financial sector crisis and restructuringof capital flows.1 in turn, problems in financial...

O C C A S I O N A L PA P E R 188

Financial Sector Crisisand Restructuring

Lessons from Asia

Carl-Johan Lindgren,Tomás J.T. Baliño, Charles Enoch,Anne-Marie Gulde, Marc Quintyn, and Leslie Teo

INTERNATIONAL MONETARY FUND

Washington DC

1999

© 1999 International Monetary Fund

Production: IMF Graphics SectionTypesetting: Joseph Ashok Kumar

Figures: In-Ok Yoon

Librar y of Congress Cataloging-in-Publication DataFinancial sector crisis and restructuring : lessons from Asia / Carl-Johan

Lindgren... [et al.].p. cm. — (Occasional paper ; 188)

Includes bibliographical references.ISBN 1-55775-871-9

1. Finance—Asia—Case studies.2. Financial crises—Asia—Case stud-ies. I. Lindgren, Carl-Johan II. Occasional paper (International Mone-tary Fund) ;188.

HC187.A2 F57 1999332’.095—dc21 99-088474

Price: US$18.00(US$15.00 to full-time faculty members and

students at universities and colleges)

Please send orders to:International Monetary Fund, Publication Services

700 19th Street, N.W., Washington, D.C. 20431, U.S.A.Tel.: (202) 623-7430 Telefax: (202) 623-7201

E-mail: [email protected]: http://www.imf.org

recycled paper

Page

Preface vii

List of Abbreviations ix

I Overview 1Origins of the Crisis 1Coping with the Crisis 5Bank Restructuring 5Cost of Restructuring 7Other Issues 7Role of the IMF 7Could the Crisis Have Been Prevented? 7

II Vulnerabilities 9Macroeconomic and Financial Weaknesses 9Structural Vulnerabilities 14

III Addressing the Emergency 16Macroeconomic Policies 16Liquidity Support 18Blanket Guarantees 18Capital Controls and Debt Rescheduling 21Immediate Closing of Financial Institutions 23

IV Monetary and Credit Policies 24Monetary Mangement 24Credit Crunch 24

V Bank Restructuring 29Broad Principles and Policies 29Institutional Arrangements 30Issues in Valuing Bank Assets 31Speed of Recapitalization 32Dealing with Troubled Institutions 33Dealing with Impaired Assets 35Cost of Restructuring 37Institutional Constraints 39Linkages to Corporate Sector Restructuring 41

VI Institutional Reform 43Prudential Supervision 43Prudential Regulation 43Regulation Governing Creditors 45

Contents

iii

CONTENTS

VII Financial Sector Reforms in IMF-Supported Programs 46

VIII Conclusions and Lessons 49

AppendicesI Indonesia 54

II Korea 67III Malaysia 79IV The Philippines 86IV Thailand 93

Bibliography 102

Boxes

SectionI 1. A Chronology of the Asian Crisis, March 1997–July 1999 2

2. Ten Critical Points in Managing and Resolving a Systemic Bank Crisis 6

II 3. Structure of the Financial System at the End of 1996 134. Weaknesses in Disclosure Practices in the Asian Crisis Countries 14

III 5. Bank Resolution Procedures: Terminology and Definitions 176. Emergency Capital Control Measures 207. Considerations Regarding the Immediate Closure of Banks in a

Systemic Crisis 218. Indonesia: Closure of 16 Banks 22

IV 9. Principal Issues in Devising a Bank Restructuring Strategy 3010. Advantages and Disadvantages of a Centralized Public Asset

Management Company 3911. The Augmented Fiscal Balance 4112. Linkage to Corporate Restructuring 42

VII 13. Financial Sector Restructuring Measures in IMF-Supported Programs47

AppendixI 14. Indonesia: Outline of Steps Toward Bank Resolution 56

II 15. Korea: KAMCO Operations 7216. Korea: Mergers and Foreign Investment in the Financial Sector 74

III 17. Malaysia: Measures Announced on March 25, 1998 83IV 18. Thailand: Capital Support Facilities 100

TablesSection

II 1. Selected Economic Indicators 102. Selected Indicators of Vulnerability 11

III 3. Liquidity Support Provided to Financial Institutions, June 1997 to June 1999 19

IV 4. Standard Deviation of Selected Monetary Indicators 255. Measures to Alleviate the Credit Slowdown 27

V 6. Summary of Measures to Address the Financial Sector Turmoil 317. Mergers, Closures, and State Interventions of Financial Institutions368. Framework for Managing Impaired Assets 379. Public Asset Management Companies in the Asian Crisis Countries 38

10. Pros and Cons of Decentralized Asset Management 4011. Authorities’Estimates for the Gross Cost of Financial

Sector Restructuring 4012. Instruments Used to Recapitalized and Purchase

Nonperforming Loans 41

iv

Contents

VI 13. Time Period for Overdue Criteria for Interest Suspension and Loan Classification 44

14. Loan Provisioning Requirements: Comparative Information 45

AppendixI 15. Indonesia: Public Cost for Financial Sector Restructuring 65

II 16. Korea: Government Ownership of Commercial Banks 7517. Korea: Public Cost for Financial Sector Restructuring 77

III 18. Malaysia: The Structure of the Malaysian Financial System 8019. Malaysia: Public Cost for Financial Sector Restructuring 85

IV 20. Philippines: Selected Banking Sector Indicators as of December 31, 1998 87

V 21. Thailand: Public Cost for Financial Sector Restructuring 101

Figures

SectionII 1. Balance on Capital and Financial Account/GDP 9

2. Real GDPGrowth 123. Real Effective Exchange Rate 124. Nominal Credit Growth to the Private Sector 12

III 5. Selected Asian Countries: Central Bank Liquidity Support 22IV 6. Indonesia: Daily Interbank Money Market Rates by Type of Bank 25

7. Growth Rate of Real Credit to the Private Sector 26VIII 8. Nominal Exchange Rate 50

AppendixI 9. Indonesia: Progress in Financial Sector Restructuring 64

II 10. Korea: Progress in Financial Sector Restructuring 76V 11. Thailand: Progress in Financial Sector Restructuring 101

v

The following symbols have been used throughout this paper:

. . . to indicate that data are not available;

— to indicate that the figure is zero or less than half the final digit shown, or that the itemdoes not exist;

– between years or months (e.g., 1994–95 or January–June) to indicate the years ormonths covered, including the beginning and ending years or months;

/ between years (e.g., 1994/95) to indicate a fiscal (financial) year.

“Billion” means a thousand million.

Minor discrepancies between constituent figures and totals are due to rounding.

The term “country,” as used in this paper, does not in all cases refer to a territorial entity thatis a state as understood by international law and practice; the term also covers some territorialentities that are not states, but for which statistical data are maintained and provided interna-tionally on a separate and independent basis.

This paper reviews the policy responses of Indonesia, Korea, and Thailand to theAsian crisis that erupted in 1997 and compares the actions of these three countrieswith those of Malaysia and the Philippines, which were buffeted by the crisis. Al -though work is still under way in all the affected countries, and thus any judgmentsare necessarily tentative, important lessons can be learned from the various experi-ences of the last two years.

The paper reflects these lessons and the emerging policies of the IMF’s Monetaryand Exchange Affairs Department (MAE). The material in this paper was originallyprepared for discussion in the IMF’s Executive Board in early September 1999. Itwas prepared under the direction of Stefan Ingves, Director of MAE, by a staff teamled by Carl-Johan Lindgren and consisting of Tomás J.T. Baliño and Charles Enoch(who have been leading the department’s work in Thailand, Korea, and Indonesia, re-spectively), Anne-Marie Gulde, Marc Quintyn, and Leslie Teo. Elena Budreckaiteand Kiran Sastry provided research assistance, and Charmane Ahmed and Janet Stan-ford secretarial support. Jeff Hayden of the External Relations Department edited thereport and coordinated its publication. The team benefited from contributions of nu-merous other MAE staff members involved in the department’s work in the fivecountries, as is acknowledged in Appendices I–V. The paper thus reflects the hands-on experience of the MAE staff and experts involved.

The paper has also benefited from comments of the IMF’s Executive Directors,colleagues in MAE, the Asia and Pacific Department (APD), and other departmentsin the IMF, as well as from colleagues in the World Bank. The views expressed in thepaper are those of the IMF staff and do not necessarily reflect the views of the na-tional authorities or IMF Executive Directors.

Preface

vii

List of Abbreviations

ix

ADB Asian Development BankBAFIA Banking and Financial Institutions Act (Malaysia)BAP Bankers Association of the PhilippinesBAPINDO Bank Pembangunan IndonesiaBBC Bangkok Bank of Commerce BCA Bank Central Asia (Indonesia)BDNI Bank Dagang Nasional IndonesiaBIBF Bangkok International Banking FacilitiesBLR Base lending rate (Malaysia)BMB Bangkok Metropolitan Bank (Thailand)BNI Bank Negara IndonesiaBRI Bank Rakyat IndonesiaBTN Bank Talringhan Nasional (Indonesia)CAMEL Capital, asset, management, equity, liquidityCDRC Corporate Debt Restructuring Committee (Malaysia)CRA Corporate Restructuring Agreement (Korea)CRCC Corporate Restructuring Coordinating Committee (Korea)DBP Development Bank of the PhilippinesFIDF Financial Institutions Development Fund (Thailand)FRAC Financial Restructuring Advisory Committee (Thailand)FRA Financial Sector Restructuring Agency (Thailand)FSC Financial Supervisory Commission (Korea)FSS Financial Supervisory Service (Korea)IBRA Indonesian Bank Restructuring AgencyKAMCO Korean Asset Management CorporationKDB Korean Development BankKDIC Korea Deposit Insurance CorporationKEXIM Korea Export-Import BankKTB Krung Thai Bank (Thailand)KTT Krung Thai Thanakit (Thailand)NIF National Investment Fund (Korea)PDIC Philippine Deposit Insurance CorporationSEC Securities and Exchange CommissionSCR Steering Committee on Restructuring (Malaysia)

Financial and corporate sector weaknesses playeda major role in the Asian crisis in 1997. These

weaknesses increased the exposure of financial insti-tutions to a variety of external threats, including de-clines in asset values, market contagion, speculativeattacks, exchange rate devaluations, and a reversalof capital flows.1 In turn, problems in financial insti-tutions and corporations worsened capital flight anddisrupted credit allocation, thereby deepening thecrisis.

As a consequence, policy responses to the crisisemphasized structural reforms in the financial andcorporate sectors in addition to the implementationof appropriate macroeconomic policies. These struc-tural measures were also necessary for macroeco-nomic policies to achieve the intended stabilization.Structural measures included dealing with nonviablefinancial institutions, establishing frameworks forrecapitalizing and strengthening viable institutions,restructuring the corporate sector, and improvingprudential regulations and supervision and marketdiscipline.2

This paper reviews the policy responses to the fi-nancial sector crisis in five Asian countries, focusingin particular on Indonesia, Korea, and Thailand. Itcomplements Lane and others (1999) and drawslessons for the future, largely based on experience inthese countries. Given that the restructuring is stillongoing, the study is necessarily selective in the is-sues it addresses and provisional in some of the an-swers it provides. Because of a combination of do-mestic and foreign factors, the crisis was particularlysevere in Indonesia, Korea, and Thailand—in thispaper referred to as the crisis countries—all ofwhich obtained the IMF’s financial support. Othercountries in the region also experienced some of theeffects of the financial turmoil. Although they did

not suffer a full-blown crisis, some of those coun-tries also adopted measures to deal with that turmoiland to strengthen their financial systems. Amongthese other countries, Malaysia and the Philippinesare useful to compare with the three crisis countries,and therefore are also discussed in this paper whenappropriate.

The structure of the paper is as follows. Section IIbriefly reviews vulnerabilities in the financial sectorin the run-up to the financial crisis. Section III dis-cusses measures taken during the initial stages of thecrisis to stabilize the system. Section IVdiscusses is-sues involved in setting monetary and credit policiesand the issue of a “credit crunch.” Section V reviewsissues related to the respective governments’strate-gies to restructure the financial sector. Section VI re-views institutional reforms undertaken to diminishthe likelihood of future financial crises. Section VIIdiscusses issues relating to IMF advice and IMF-supported programs. Conclusions and lessons arepresented in Section VIII. Appendices I–Vpresentcase studies from Indonesia, Korea, Malaysia, thePhilippines, and Thailand. The studies discuss in de-tail the financial sector problems and the steps takento address them. They set the stage for the compar-isons and lessons that are drawn in the main body ofthe paper.

The following paragraphs offer a brief analysis ofthe crisis and summarize the paper’s main findings.

Origins of the Crisis

Financial and corporate sector weaknesses com-bined with macroeconomic vulnerabilities to sparkthe crisis (see Box 1 for a chronology of events).Formal and informal currency pegs, which discour-aged lenders and borrowers from hedging, alsocontributed to the outbreak. Capital inflows hadhelped fuel rapid credit expansion, which loweredthe quality of credit and led to asset price inflation.The inflated asset prices encouraged further capitalinflows and lending, often by weakly supervisednonbank financial institutions. Highly leveragedcorporate sectors, especially in Korea and Thailand,

I Overview

1

1For simplicity, “bank” and “financial institution” are used in-terchangeably in this paper. When referring to a specific type offinancial institution (e.g., commercial bank, merchant bank), thatreference will be used in full.

2Nonviable financial institutions are those judged unable tomaintain minimum thresholds of liquidity, solvency, andprofitability.

I OVERVIEW

2

Box 1. A Chronology of the Asian Crisis, March 1997–July 1999

1997

March 3 Thailand First official announcement of problems in two unnamed financecompanies, and a recapitalization program.

March–June Thailand Sixty-six finance companies secretly receive substantial liquiditysupport from the Bank of Thailand. Significant capital outflows.

April Malaysia Bank Negara Malaysia imposes limits on bank lending to the prop-erty sector and for the purchase of stocks.

June 29 Thailand Operations of 16 finance companies suspended and a guarantee ofdepositors’and creditors’funds in remaining finance companies an-nounced.

July 2 Thailand Baht is floated and depreciates by 15–20 percent.

Early July Indonesia Pressure on the rupiah develops.

July 8–14 Malaysia Bank Negara Malaysia intervenes aggressively to defend the ringgit:efforts to support the ringgit are abandoned; ringgit is allowed tofloat.

July 11 Indonesia Widening of the rupiah’s band.

July 11 Philippines Peso is allowed more flexibility.

July 13 Korea Several Korean banks are placed on negative credit outlook by ratingagencies.

July 24 All “Currency meltdown”—severe pressure on rupiah, baht, ringgit, andpeso.

August 5 Thailand Measures adopted to strengthen financial sector. Operations of 42 fi-nance companies suspended.

August 14 Indonesia Authorities abolish band for rupiah, which plunges immediately.

August 20 Thailand Three-year Stand-By Arrangement with IMF approved.

August 25 Korea Government guarantees banks’external liabilities; withdrawal ofcredit lines continues.

October 14 Thailand Financial sector restructuring strategy announced; Financial SectorRestructuring Agency and asset management company established;blanket guarantee strengthened; new powers to intervene in banks.

October 24 Thailand Emergency decrees to facilitate financial sector restructuring.

October 31 Indonesia Bank resolution package announced; 16 commercial banks closed;limited deposit insurance for depositors in other banks; other bankclosures to follow.

November 5 Indonesia Three-year Stand-By Arrangement with IMF approved.

November 19 Korea Exchange rate band widened. Won falls sharply.

Mid-November Thailand Change in government. Significant strengthening of economic reformprogram.

November Korea Korea Asset Management Corporation’s (KAMCO) nonperformingasset fund is established.

December 4 Korea IMF approves three-year Stand-By Arrangement but rollover ofshort-term debt continues to decline.

December 8 Thailand Fifty-six suspended finance companies are permanently closed.

Mid-December Indonesia Deposit runs on banks, accounting for half of banking system assets.

December 18 Korea New government is elected; commitment to program is strengthened.

Origins of the Crisis

3

December 24 Korea Foreign private bank creditors agree to maintain exposure temporarily.

December 29 Korea Legislation passed strengthening independence for Bank of Koreaand creating Financial Supervision Commission.

December 31 Thailand Bank of Thailand intervention in a commercial bank; shareholders’stakes eliminated.

December Korea Fourteen merchant banks are suspended and two large commercialbanks taken over by the government.

1998

January 1 Malaysia Measures announced to strengthen prudential regulations.

January 15 Indonesia Second IMF-supported program announced. Indonesian Bank Re-structuring Agency (IBRA) established and blanket guarantee an-nounced.

January 20 Malaysia Bank Negara Malaysia announces blanket guarantee for all depositors.

January 23 Thailand Bank of Thailand intervenes in two commercial banks; shareholderseliminated.

January 26 Indonesia Indonesian Bank Restructuring Agency (IBRA) established and blan-ket guarantee announced.

January 28 Korea Agreement with external private creditors on rescheduling of short-term debt.

January Korea Ten of 14 suspended merchant banks closed; 20 remaining merchantbanks are required to submit rehabilitation plans.

February 15 Korea New president and government take office.

February Indonesia President Suharto reelected. Doubts about future of financial sectorprogram grow stronger amid political uncertainty. Rupiah depreciatesfurther and currency board is debated.

March 11 Thailand One commercial bank purchased by foreign strategic investor.

March 25 Malaysia Program to consolidate finance companies and to recapitalize com-mercial banks is announced.

March 31 Thailand New loan classification and loss provisioning rules introduced.

March Philippines Three-year Stand-By Arrangement agreed with IMF.

April 4 Indonesia IBRA closes seven banks and takes over seven others.

End of April Korea Four of 20 merchant banks’rehabilitation plans rejected; banks areclosed.

May 18 Thailand Bank of Thailand intervention in seven finance companies; share-holders eliminated.

Mid-May Indonesia Widespread riots. Rupiah depreciates, deposit runs intensify, andBank Indonesia must provide liquidity.

May 21 Indonesia President Suharto steps down.

May 29 Indonesia A major private bank taken over by IBRA.

June 5 Indonesia International lenders and Indonesian companies agree on corporatedebt rescheduling.

June 29 Korea For the first time, government closes commercial banks (five smallones). Two merchant banks are closed and two merged with commer-cial banks.

June 30 Korea New loan classification and loss provisioning rules introduced.

I OVERVIEW

and large unhedged short-term debt made the crisiscountries vulnerable to changes in market sentimentin general and exchange and interest rate changes inparticular. Malaysia and the Philippines were lessvulnerable.

Weaknesses in bank and corporate governanceand lack of market discipline allowed excessive risktaking, as prudential regulations were weak orpoorly enforced. Close relationships between gov-ernments, financial institutions, and borrowers wors-ened the problems, particularly in Indonesia andKorea. More generally, weak accounting standards,especially for loan valuation, and disclosure prac-tices helped hide the growing weaknesses from poli-

cymakers, supervisors, market participants, and in-ternational financial institutions—while those indi-cators of trouble that were available seem to havebeen largely ignored. In addition, inadequacies in as-sessing country risk on the part of the lenders con-tributed to the crisis.

The crisis was triggered by the floating of the Thaibaht in July 1997. Changing expectations led to thedepreciation of most other currencies in the region,bank runs and rapid withdrawals of foreign privatecapital, and dramatic economic downturns. Whenthe crisis broke, Indonesia, Korea, and Thailand re-quested IMF assistance, both to obtain financial sup-port and to restore confidence.

4

Box 1 (Concluded)

June Malaysia Danaharta, an asset management company, is established.

Mid-July Indonesia Authorities allow market-determined interest rates on Bank Indone-sia bills.

August 14 Thailand Comprehensive financial sector restructuring plan announced, in-cluding facilities for public support of bank recapitalization. Inter-vention in two banks and five finance companies; shareholders’stakes eliminated.

August 30 Thailand Majority ownership in one medium-sized commercial bank by for-eign strategic investor.

August Malaysia Danamodal (bank restructuring and recapitalization agency) isestablished.

September 23 Indonesia Indonesia’s bilateral external debt to official creditors refinanced.

September 30 Indonesia Bank Mandiri created through merger of four largest state-ownedbanks. Plans announced for joint government-private sector recapital-ization of private banks.

September Malaysia Capital controls introduced, exchange rate pegged, disclosure re-quirements relaxed, and measures to stimulate bank lending adopted.

October 6 Indonesia Amended Banking Law passed, which included strengthening ofIBRA.

1999

February 15 Malaysia Capital controls replaced with declining exit levies.

March 13 Indonesia Government closes 38 banks and IBRAtakes over seven others. Eli-gibility of nine banks for joint recapitalization with government an-nounced.

April 21 Indonesia Closure of one joint-venture bank.

April Indonesia Government announces a plan to recapitalize the three other statebanks (all insolvent).

June 30 Indonesia Eight private banks recapitalized jointly through public and privatefunds.

July 5 Indonesia Government announces plan for resolution of IBRAbanks.

July 31 Indonesia Legal merger of component banks of Bank Mandiri.

July Thailand One small private bank intervened and put up for sale; one majorbank announces establishment of an asset management company.

Bank Restructuring

Coping with the Crisis

The initial priorities in dealing with the crisis wereto stabilize the financial system and to restore confi-dence in economic management. Forceful measureswere needed to stop bank runs, protect the paymentsystem, limit central bank liquidity support, mini-mize disruptions to credit flows, maintain monetarycontrol, and stem capital outflows. In the crisiscountries, emergency measures, such as the intro-duction of blanket guarantees and bank closings,were accompanied by comprehensive bank restruc-turing programs and supported by macroeconomicstabilization policies.

Closings of the most insolvent or nonviable finan-cial institutions were used initially to stem rapidlyaccumulating losses and central bank liquidity sup-port. However, the experience of Indonesia showedthat in a systemic crisis bank closings can lead toruns on other banks, if not accompanied by properinformation, strong overall economic management,and a blanket guarantee.

Blanket guarantees for depositors and creditorswere used in the crisis countries and in Malaysia torestore confidence and to protect banks’funding.Despite the enormous contingent costs and moralhazard problems involved, governments consideredguarantees preferable to collapses of their bankingsystems. The guarantees were effective in stabilizingbanks’domestic funding—although in some cases ittook some time to gain credibility—but were less ef-fective in stabilizing banks’foreign funding (Korearesponded with voluntary debt rescheduling andMalaysia adopted capital controls). In Indonesia, ablanket guarantee was introduced only after an at-tempt to use a limited guarantee had backfired.

Liquidity support by central banks was reducedafter the closure of the weakest financial institutionsand the introduction of the blanket guarantees. Mon-etary control was maintained through sterilizationmeasures—offsetting sales or purchases of securitiesby the central bank—in all countries, except initiallyin Indonesia. Monetary policy in all countries fo-cused on the exchange rate, short-term interest rates,and the level of international reserves, rather than ontraditional monetary aggregates, which had becomeunstable.

Credit growth slowed as demand contracted andsupply plummeted, with bankers becoming more se-lective in their lending behavior. A heightened per-ception of credit risk, funding constraints, and aweakening capital position further constrainedcredit. In such circumstances, direct or indirect mea-sures to stimulate new credit are unlikely to be suffi -cient to restore normal lending: that will take a re-turn of profitability and solvency in the banking andcorporate sectors.

Bank Restructuring

Comprehensive bank restructuring strategies in thecrisis countries and in Malaysia sought to restore fi-nancial sector soundness as soon as possible, and atleast cost to the government, while providing an ap-propriate incentive structure for the restructuring.(See Box 2 for a list of critical steps in resolving asystemic banking crisis.) The strategies includedsetting up appropriate institutional frameworks, re-moving nonviable institutions from the system,strengthening viable institutions, dealing with value-impaired assets, improving prudential regulationsand banking supervision, and promoting trans-parency in financial market operations.

Key principles for bank restructuring strategies inthe crisis countries have been the application of uni-form criteria to identify viable and nonviable institu-tions, removal of existing owners from insolvent in-stitutions, and encouragement of new private capitalcontributions, including from the foreign sector.Public support has sought to complement privatesector contributions; liberalization of foreign owner-ship rules encouraged foreign participation.

Strategies must be adapted to fit countries’cir-cumstances. Systemic bank restructuring is a com-plex medium-term process that requires carefultailoring. Accordingly, while the broad componentsof the restructuring strategies were similar, imple-mentation details differed across countries accord-ing to the precise nature of the problems, legal andinstitutional constraints, and each government’spreferences.

Valuation of bank assets is crucial for determin-ing bank viability but is very difficult in a crisis en-vironment, as markets are thin and values shift withchanging circumstances. Tighter rules for loan clas-sification, loss provisioning, and interest suspen-sion were introduced to guide the valuation process.Dif ferent valuation procedures, including by banksthemselves, external or international auditors, orsupervisors, were used to provide the authoritieswith the best available data. Regardless of dataquality, decisions had to be made as much as possi-ble on the basis of uniform and fully transparentcriteria.

Strengthening viable institutions involved assetvaluation, loss recognition, and recapitalization.When banks breached minimum capital adequacyrequirements, recapitalization and rehabilitation be-came mandatory, often under binding memoranda ofunderstanding with the supervisory authorities. Inthe crisis countries loan-loss provisioning rules orcapital adequacy requirements were implementedgradually—but transparently—to give banks time torestructure and mobilize new capital and to avoidaggravating credit supply problems. Public sector

5

I OVERVIEW

equity support was also provided to viable banks,subject to stringent conditions.

The authorities intervened in institutions thatfailed to raise capital and faced insolvency throughsuch techniques as government recapitalization/na-tionalization, mergers, sales, use of bridge banks andasset management companies, purchase and assump-tion operations, and liquidation. Shareholders typi-cally absorbed losses until their capital was fullywritten off. In all the countries, the governments aimat reprivatizing the nationalized financial institutionsas soon as possible; in this, Korea and Thailand havealready made significant progress.

Management of impaired assets, including non-performing loans, is one of the most complex partsof financial restructuring. Impaired assets can eitherbe dealt with by the financial institutions them-selves, by bank-specific or centralized asset manage-ment companies, or under liquidation procedures.Speed of disposal has to be considered. Assets haveto be managed and disposed so as to preserve valuesand maximize recovery, while at the same time cre-ate the right incentives so as not to undermine bor-rower discipline throughout the system. The choiceof asset management structure should depend on the

nature of the asset and available management capa-bilities. Nonperforming loans with reasonablechances of recovery are generally better managed inbanks.

A centralized asset management company typi-cally involves government ownership, comparedwith decentralized asset management companies,which tend to be privately owned and bank specific.All asset management companies seek to providebetter management structures for problem assets andto relieve banks’balance sheets. Asset sales bybanks to asset management companies should notamount to back-door capitalization of banks (andbailout of shareholders) by receiving inflated prices,a matter complicated by the above-mentioned diffi -culty of valuing impaired assets. Because bankshave to take a loss when they sell loans to an assetmanagement company (public or private), capitalscarcity may limit their capacity to do so. Indonesia,Korea, and Malaysia have opted for a centralizedpublic asset management company, while Thailandestablished a public asset management company thatonly deals with residual assets of closed financecompanies, and has encouraged the establishment ofbank-specific asset management companies.

6

The sequence presented below describes the differentphases one encounters when dealing with a major sys-temic financial sector crisis. This sequence is based onthe assumption that a country’s financial sector haspublic good aspects and, hence, that solving such a cri-sis warrants substantial public sector involvement. Thedifferent steps, from origin through recognition andresolution, and preventive measures are discussed inthis paper. Although specific actions may differ amongcountries based on the depth of the crisis, the composi-tion of the financial sector before the crisis, local cir-cumstances and preferences, and the contents and se-quence of the basic building blocks and strategies aresimilar across countries.

Steps 1–4. The acute crisis phase: measures to stop thepanic and stabilize the system.

1. The crisis usually begins because, in one form or an-other, there is excessive leverage in the economy. In theearly stages there may also be a degree of denial on thepart of the banks and the government.

2. Bank runs by creditors and depositors start and in-tensify. The central bank responds by providing liquid-ity support to the affected banks.

3. When central bank liquidity is unable to stop theruns, the government announces a blanket guaranteefor depositors and creditors. Such a measure is intended

to reduce uncertainty and to allow time for the govern-ment to begin an orderly restructuring process.

4. All along, the central bank tries to sterilize its liq-uidity support to avoid a loss of monetary control.

Steps 5–8.The stabilization phase: measures to re-structure the system.

5. The authorities design the tools needed for a com-prehensive restructuring, including the required legal,financial, and institutional framework.

6. Losses in individual institutions are recognized. Theauthorities shift the focus from liquidity support to sol-vency support.

7. The authorities design a financial sector restructur-ing strategy, based on a vision for the postcrisis struc-ture of the sector.

8. Viable banks are recapitalized, bad assets are dealtwith, and prudential supervision and regulations aretightened.

Steps 9–10.The recovery phase: measures to normal-ize the system.

9. Nationalized banks are reprivatized, corporate debtis restructured, and bad assets are sold.

10. The blanket guarantee is revoked, which, if prop-erly handled, is a nonevent because the banking systemhas been recapitalized and is healthy again.

Box 2. Ten Critical Points in Managing and Resolving a Systemic Bank Crisis

Could the Crisis Have Been Prevented?

Cost of Restructuring

The gross costs of the bank restructurings aremassive. Estimates put the public sector costs in thethree crisis countries and Malaysia between 15 and45 percent of GDP. The estimates may increase iffurther losses are uncovered, but they may also dropdepending upon the proceeds from asset sales andprivatization. The revenue generated by these saleswill not be known for several years. There are, in ad-dition, efficiency gains and wealth effects resultingfrom the restructuring. Initially, the costs weremainly carried by the central banks in the form ofliquidity support to ailing banks. Only recently havegovernments started to refinance this liquidity by is-suing domestic government bonds.

The fiscal implications of the crisis were esti-mated by imputing the carrying costs of the debt cre-ated to finance the restructuring. Full and transparentrecording in the fiscal accounts of all costs incurredby the government, including capital costs, is impor-tant for fiscal analysis. The very large costs of thecrisis may affect medium-term fiscal sustainability.

Other Issues

Government “ownership” of the reform programsand strong leadership are necessary to take charge ofand implement the complex microeconomicprocesses that a systemic bank restructuring entail.In the crisis countries, political changes had a posi-tive impact on the pace and resolve of the restructur-ing process. Only domestic constituencies can dealwith the legal and institutional factors that are pre-requisites for success, but that also can bring theprocess to a halt. Restructuring has to take into ac-count human resource constraints and legal issues,given that it typically has major effects on privatewealth.

Corporate sector problems represent the flip sideof banks’nonperforming loans. Bank restructuringshould be accompanied by corporate debt restructur-ing, which has been lagging and is now delaying thebank restructuring process. At the same time, finan-cial sector restructuring should be given priority asthe governance structure of banks and their pruden-tial framework provide powerful levers to bringabout the corporate restructuring reform.

Prudential regulation and supervision have beenstrengthened to foster better bank governance andstronger market discipline. In all the countries, do-mestic standards are being brought closer to interna-tional best practices, including areas such as foreignexchange exposure, liquidity management, con-nected lending, loan concentration, loan provision-ing, data disclosure, and qualifications for owners

and managers. Steps have been taken to strengthenthe autonomy and authority of supervisors, upgradetheir powers and skills, and improve on-site exami-nation, off-site monitoring, and analysis techniques.

Role of the IMF

The IMF-supported programs in Indonesia, Korea,and Thailand centered on financial sector reform, notonly because financial sector problems were a rootcause of the crisis but also because reestablishingbanking system soundness was crucial for restoringmacroeconomic stability. Although the IMF was ableto draw on both its past experience and its analyticalwork, the specific circumstances of each countryadded dimensions that required careful tailoring ofthe reform and resolution strategies for each country,often taking into account the authorities’sometimesstrong preferences. The design of the reform strate-gies required access to bank-by-bank supervisorydata, which was provided in the crisis countries.

Letters of Intent and Memoranda of Economic andFinancial Policies laid out the strategies and sequenc-ing. The IMF-supported programs required a delicatebalance between the needs for short-term IMF condi-tionality and the medium-term nature of financialsector restructuring, which often involves steps andnegotiations beyond the authorities’direct control.

Cooperation with the World Bank and other inter-national organizations was close from, or soon after,the beginning, with somewhat different divisions oflabor in each country. The IMF took the lead in as-sisting the authorities in designing the overall restruc-turing program of the three crisis countries, while theWorld Bank took charge of specific areas of programformulation and implementation. Most tasks havebeen done jointly to provide the authorities with thebest possible advice and to use the resources of thetwo institutions as efficiently as possible.

Could the Crisis Have Been Prevented?

More transparency in macro- and microeconomicdata and policies would have exposed vulnerabilitiesearlier and helped lessen the crisis. Better regulatoryand supervisory frameworks would have helped, butsupervisors would most likely not have been able totake necessary actions in the middle of the economicboom. No one foresaw the sudden massive erosionof loan values, once market sentiment changed andexchange rates collapsed.

Broad-based reforms are under way to strengthenthe institutional, administrative, and legal frame-works in the crisis countries, based on evolving in-

7

I OVERVIEW

ternational best practices, codes, core principles, andstandards. The crisis has shown the need to tailorprudential policies so that resilience is built up intimes of economic booms to deal more easily withinevitable economic downturns.

International efforts have been undertaken to re-duce the likelihood and intensity of future crises. Ini-

tiatives include work on the international financialarchitecture, the Financial Stability Forum, andfinancial sector stability assessments. The BaselCommittee on Banking Supervision has formulatedimprovements to regulation and supervision of inter-national lenders to address weaknesses that con-tributed to the Asian crisis.

8

The Asian financial crisis involved several mutu-ally reinforcing events, starting with the devalu-

ation of the Thai baht in July 1997, and followed bydevaluations of other currencies, the attack on theHong Kong dollar in October 1997, a rapid with-drawal of foreign private capital, bank runs, sover-eign downgrades, and a dramatic decline in real eco-nomic activity.3 A combination of financial systemand corporate sector vulnerabilities and weaknessescontributed to the crises and magnified the negativeimpact of exchange rate devaluations and foreigncapital withdrawals on financial institutions.Thissection highlights some of these vulnerabilities,which were present in all the crisis countries, albeitdiffering in specific aspects.

Macroeconomic and Financial Weaknesses

A key vulnerability arose from the large capital in-flows—especially those deriving from foreign bor-rowing. These inflows were equivalent to 3.5 per-cent of GDPannually in Indonesia, 2.5 percent inKorea, and 10 percent in Thailand during 1990–96(Figure 1). They were encouraged by high economicgrowth, low inflation, and relatively healthy fiscalperformance (Tables 1 and 2, and Figure 2), finan-cial sector and capital account liberalization, integra-tion into global capital markets, formal or informalexchange rate pegs (Figure 3), and various incen-tives created by the government.4 Capital flows alsoreflected conditions in the global financial system,

including low interest rates and weaknesses in riskmanagement by lenders in industrialized countries.The bulk of these inflows reflected direct borrowingby banks (Korea and Thailand) and corporations (In-donesia): this was especially evident in Thailandright before the crisis.5 In contrast, in Malaysia, in-flows of foreign direct investment were larger thandirect borrowing and portfolio inflows, while capitalinflows in the Philippines (particularly portfolio in-flows) had only recently become significant.

Inflexible exchange rate regimes complicatedmacroeconomic management and increased vulner-ability. The nominal exchange rate had depreciatedin a predictable manner in Indonesia, and was

II Vulnerabilities

9

3This paper does not address the causes of the crisis. Detailedexpositions on this subject can be found in Lane and others(1999), International Monetary Fund (1997), and InternationalMonetary Fund (1998). Other studies, such as Furman andStiglitz (1998), Goldstein (1998), and Radelet and Sachs (1998),have also addressed the subject.

4For example, in Thailand, bank lending and borrowingthrough Bangkok International Banking Facilities received favor-able tax treatment, while in the Philippines, banks were subject tolower taxes on onshore income from foreign currency loans com-pared to that from domestic currency loans.

5In Thailand other inflows were 8 percent of GDPin 1996compared to 3 percent of GDPfor direct investment and portfolioinflows; in Indonesia and Korea, other inflows were on average 1percent of GDP, and direct and portfolio investment 4 percent.

Indonesia ThailandMalaysiaPhilippines

Korea

–20

–15

–10

–5

0

5

10

15

98979695949392911990

Source: IMF, World Economic Outlook.

Figure 1. Balance on Capital and FinancialAccount/GDP(In percent)

closely linked to the U.S. dollar (or a basket of cur-rencies) in Korea, Malaysia, the Philippines, andThailand. The broadly stable exchange rate createdincentives for borrowing in foreign exchange asborrowers underestimated the risks associated withforeign currency exposure.6 Lenders, meanwhile,ignored the fact that lending in foreign exchange in-volved substantial credit risk. Maturity mismatchesin banks’portfolios, and currency mismatches on

corporations’balance sheets aggravated the prob-lem. A long history of stable exchange rates alsoundermined incentives to introduce adequate pru-dential rules on, and monitoring of, foreign cur-rency exposures. The three crisis countries were es-pecially vulnerable to capital outflows andexchange rate devaluations because of the signifi-cant amount of short-term foreign currency debt,which was mostly unhedged. Furthermore, thegrowth of this debt outpaced growth in usable for-eign exchange reserves during most of the 1990s,making these countries increasingly susceptible to adeterioration in market sentiment and large capital

10

II VULNERABILITIES

Table 1. Selected Economic Indicators(In percent or ratios)

1990 1991 1992 1993 1994 1995 1996 1997 1998 19991

IndonesiaReal GDP growth 9.0 8.9 7.2 7.3 7.5 8.2 8.0 4.6 –13.6 –3.9Inflation 7.8 9.4 7.5 9.7 8.5 9.4 7.9 6.6 60.7 25.4Current account balance/GDP –2.8 –3.4 –2.2 –1.5 –1.7 –3.3 –3.2 –3.0 –0.1 2.8Central government balance/GDP 1.34 0.04 –1.15 –0.71 0.01 0.77 1.16 –0.67 –4.46 –6.48Broad money growth 29.7 24.6 22.6 21.1 21.8 26.7 27.0 27.4 61.7 15.6Private sector credit/GDP 46.1 45.8 45.5 48.9 51.9 53.5 55.4 61.0 51.6 . . .

KoreaReal GDP growth 9.5 9.1 5.1 5.8 8.6 8.9 7.1 5.5 –5.5 2.0Inflation 8.6 9.3 6.2 4.8 6.3 4.5 4.9 4.4 7.5 1.8Current account balance/GDP –0.8 –2.8 –1.3 0.3 –1.0 –1.9 –4.7 –1.8 13.1 7.1Central government balance/GDP –0.67 –1.62 –0.49 0.64 0.32 0.35 0.28 0.28 –3.78 –5.12Broad money growth 17.2 21.9 14.9 16.6 18.7 15.6 15.8 14.1 25.2 . . .Private sector credit/GDP 52.5 52.8 53.3 54.2 56.8 57.0 61.8 69.8 73.6 . . .

MalaysiaReal GDP growth 9.6 8.6 7.8 8.3 9.3 9.4 8.6 7.7 –7.5 –1.6Inflation 2.8 2.6 4.7 3.5 3.7 3.4 3.5 2.7 5.3 3.8Current account balance/GDP –2.1 –8.8 –3.8 –4.8 –7.8 –10.0 –4.9 –5.1 12.3 8.7Central government balance/GDP –3.08 –2.48 0.13 0.52 1.45 1.30 1.07 2.58 –1.91 –6.05Broad money growth 18.2 24.4 18.1 23.8 15.8 18.2 23.7 9.6 1.3 6.1Private sector credit/GDP 71.4 75.3 74.3 74.1 74.6 84.8 89.8 100.4 108.7 . . .

PhilippinesReal GDP growth 3.0 –0.6 0.3 2.1 4.4 4.7 5.8 5.2 –0.5 2.3Inflation 14.1 18.7 9.0 7.6 9.1 8.1 8.4 6.0 9.7 8.5Current account balance/GNP –5.8 –1.9 –1.6 –5.5 –4.6 –4.3 –4.4 –5.1 1.8 2.1Central government balance/GNP –3.80 –2.40 –1.30 –1.60 –1.70 –1.30 –0.60 –0.70 –2.60 –2.70Broad money growth 15.5 15.5 11.0 24.6 26.5 25.3 15.8 20.9 7.4 15.0Private sector credit/GNP 20.5 18.9 21.5 27.2 30.0 38.2 50.0 57.6 50.5 46.9

ThailandReal GDP growth 11.6 8.1 8.2 8.5 8.6 8.8 5.5 –0.4 –8.0 1.0Inflation 6.0 5.7 4.1 3.4 5.1 5.8 5.9 5.6 8.1 0.5Current account balance/GDP –8.3 –7.5 –5.5 –5.0 –5.4 –7.9 –7.9 –1.9 12.2 8.8Central government balance/GDP 4.60 4.14 2.53 1.98 1.98 2.49 1.04 –1.62 –2.88 –3.84Broad money growth 26.7 19.8 15.6 18.4 12.9 17.0 12.6 16.4 9.5 4.7Private sector credit/GDP 64.5 67.7 72.2 79.8 90.9 97.5 100.0 116.3 109.5 . . .

Sources: IMF, International Financial Statistics;World Economic Outlook; and national authorities.11999 IMF estimates.

6In addition, domestic nominal interest rates were above for-eign rates, especially with regard to yen rates.


outflows.7 In addition, material adverse changeclauses in debt contracts shortened the effectivematurity of long-term debt, increasing vulnerabilityto negative events.8 Malaysia was less vulnerablebecause foreign currency borrowing was lower dueto a requirement of official approval above a certainlimit.

The capital inflows helped fuel rapid credit expan-sion that led to strains—asset price inflation and ex-cessive risk taking—which increased the vulnerabil-ity of the financial systems. In Korea, Malaysia, andThailand private sector credit in nominal terms ex-panded rapidly during the 1990s, at an average rate

of 15 to 20 percent compared to inflation rates of 3to 10 percent (Figure 4). Total commercial bank andnear-bank assets grew from between 50 and 100 per-cent of GDPin 1992 to between 150 and 200 percentof GDPat the end of 1996 (Box 3). As a comparison,deposit money banks held assets equal to 30 percentof GDPin Mexico, 48 percent in Brazil, 80 percentin the United States, 136 percent in the EuropeanUnion, and 300 percent in Japan.9 The Asianeconomies were in a self-reinforcing cycle—growthin credit reinforced the investment booms, which inturn encouraged further capital inflows and lending.This growth also led to asset price inflation (espe-cially in Malaysia and Thailand), which encouragedlending to the real estate sector and inflated collat-eral values. Meanwhile, banks were increasingly ex-posed to credit and foreign exchange risks and tomaturity mismatches, to the extent that foreign bor-rowing was short term and domestic lending longterm, thus increasing the countries’vulnerability to

11

7By mid-1997, total outstanding claims held by foreign bankson domestic residents in the three crisis countries amounted to$232 billion, of which $151 billion was short term. Short-termdebt amounted to 20 percent of total foreign debt in Indonesia, 44percent in Korea, 50 percent in Malaysia, 60 percent in the Philip-pines, and 30 percent in Thailand. In Indonesia, Korea, and Thai-land, the ratio of short-term liabilities to international reserveswas above 1; in Malaysia it was 0.6, and in the Philippines it was0.8 (see also Lane and others, 1999).

8In some cases, such clauses would permit the lender to requireimmediate repayment if a country’s bond or sovereign ratingwere downgraded.

Table 2. Selected Indicators of Vulnerability(Period ended December 1996)

Indonesia Korea Malaysia Philippines Thailand

Macro indicatorsInflation >5% • • •Fiscal deficit >2% of GDPPublic debt >50% of GDP •Current account deficit >5% of GDP •Short-term flows >50% current account deficit1 • • • • •Capital inflows >5% of GDP • • • •Ratio of short-term debt to international reserves >1 2 • • •

Financial sector indicatorsRecent financial sector liberalization • • • •Recent capital account liberalization •Credit to the private sector >100% of GDP • • •Credit to the private sector, real growth >20% • •Emphasis on collateral when making loans • • • • •Estimated share of bank lending to the real estate sector >20%3 • • • •Stock of nonperforming loans >10% of total loansStock market capitalization (as percent of GDP) 40 30 310 98 56

Source: IMF, International Financial Statistics; World Economic Outlook;World Bank and IFC.Note:The cutoff points are based on the relevant literature that attempts to predict currency and banking crises (Kaminsky, Lizondo, and Reinhart,

1997, for currency crisis; and Hardy and Pazarbas ıog lu, 1998, for banking crisis).1Defined as the sum of net portfolio and other investments in the financial accounts.2As of June 1997.3At the end of 1997. Includes indirect exposure through collateral.

9Data for the end of 1998 for the European Union and the endof 1996 for the other countries. Data for nonbank financial insti-tutions in these countries are not readily available.

II VULNERABILITIES

outflows.10 Rapid growth also strained banks’capac-ity to assess risk adequately.

In most countries, the growing nonbank financialinstitutions held riskier assets and more volatile fi-nancing than commercial banks, which made themincreasingly vulnerable to a decline in asset qualityand to a change in investor and depositor senti-ment.11 Nonbank financial institutions had grownvery rapidly in recent years; they were favored bythe easier licensing requirements (Thailand) and lessstringent regulations, including lower capital re-quirements (Korea and the Philippines) than thoseapplied to commercial banks.12 Merchant banks inKorea and finance companies in Thailand were thefirst institutions to face liquidity shortfalls, andmany became insolvent and had to be closed.13

The corporate sector in Korea and Thailand washighly leveraged, a factor that, in combination withthe pervasive nature of the corporate crisis, signifi-cantly deepened the banking crisis. Average debt toequity ratios of listed companies were around 400percent in both countries at the end of 1996. By con-trast, ratios in Indonesia, Malaysia, and the Philip-

12

10Banks had lent substantial amounts in foreign currency toborrowers without secure foreign exchange revenue streams. Thecorporate sector’s repayment capacity became severely impairedonce the currencies started to depreciate, leading to corporate in-solvencies and major problems for the banks.

11Nonbank financial institutions had become increasingly im-portant compared to commercial banks in Korea and Thailand.This trend has been particularly striking in Korea, where com-mercial banks’share of total deposits has fallen from 71 percentin 1980 to 30 percent at the end of 1996 to the benefit of invest-ment trust companies, insurance companies, and other nonbankfinancial institutions.

12There is an argument for less stringent prudential require-ments for nonbank financial institutions, insofar as they performa narrower range of activities. However, these institutions in theAsian crisis countries operated broadly like commercial banks.

13Finance companies in Malaysia also faced liquidity short-falls. There, the government’s policy has been to strengthen thesector through mergers (see Section V).

Indonesia ThailandMalaysia

PhilippinesKorea

–15

–10

–5

0

5

10

15

9998979695949392911990

Source: IMF, World Economic Outlook.Note: Data for 1999 are IMF staff estimates.

Figure 2. Real GDP Growth(In percent)

IndonesiaThailand

Malaysia PhilippinesKorea

1993 1994 1995 1996 1997 1998 19990

20

40

60

80

100

120

140

Source: IMF, International Financial Statistics.

Figure 3. Real Effective Exchange Rate(June 1997 = 100)

Indonesia

ThailandMalaysia PhilippinesKorea

9392911990 94 95 96 97 98–20

0

20

40

60

80


Figure 4. Nominal Credit Growth to thePrivate Sector(In percent)


13

Commercial banks dominated the financial systemin Indonesia. Out of a total of 238 commercial banks,there were 7 state-owned banks, 27 regional govern-ment banks, 160 private banks, 34 joint-venturebanks, and 10 foreign banks. In addition, there wereapproximately 9,200 rural banks. Nonbank financialinstitutions included 252 finance companies, 163insurance companies, about 300 pension and provi-dent funds, and 39 mutual fund companies. Totalassets of the system were equivalent to about 90 per-cent of GDP. Commercial banks held 84 percent oftotal assets while rural banks held about 2 percent.The remaining assets were held by finance compa-nies (7 percent of total assets), insurance companies(5 percent), and other nonbank financial institutions(2 percent).

Korea’s financial system was the most differentiatedamong these countries, with 26 commercial banks, 52branches of foreign commercial banks, 7 specializedand development banks, 30 merchant banks, 30investment trust companies, 33 life insurance compa-nies, 17 nonlife insurance companies, 56 securitiescompanies, 19 investment advisory companies, 230mutual savings companies, 2 credit guarantee funds,and approximately 6,000 credit unions, mutual creditfacilities, and community credit cooperatives. Totalassets of the system were close to 300 percent of GDP.Commercial banks alone accounted for 52 percent oftotal assets, while specialized and development banksaccounted for 17 percent.1 Merchant banks held a fur-ther 5 percent; insurance companies 7 percent; whilethe remaining 19 percent were held by other types ofinstitutions.

In Malaysia, the financial system included 35 com-mercial banks, 39 finance companies, 12 merchantbanks, 7 discount houses, 4 pension and providentfunds, 62 insurance companies, 6 unit trusts, 7 develop-ment institutions, and a savings bank. Total assets ofthe system were equivalent to 300 percent of GDP.Commercial banks accounted for 70 percent of total as-sets of the banking system (comprising the commercialbanks, finance companies, and merchant banks), andmerchant banks and finance companies for 30 percentof total assets.

Although the financial sector in the Philippines in-cluded different types of banks and nonbanks, the sec-tor was dominated by commercial banks. At the end of

1996, there were 49 commercial banks, 3 specializedgovernment banks, 109 thrift banks, approximately 800rural banks, 129 insurance companies, 12 nonbank fi-nancial institutions with quasi-banking functions, and alarge number of small nonbank institutions withoutquasi-banking functions. The number of foreign banksthat operate wholly owned branches in the Philippinesis currently capped at 14. Total assets of the systemwere equivalent to 115 percent of GDP. Commercialbanks held 82 percent of total assets, thrift banks ac-counted for 9 percent, rural banks 1 percent, and non-banks the remaining 8 percent.

In Thailand, the financial system included 29 com-mercial banks, 91 finance and securities companies, 7specialized state-owned banks, approximately 4,000savings and agricultural cooperatives, 15 insurancecompanies, 880 private provident funds, and 8 mutualfund management companies. Out of the 29 commer-cial banks, 14 were branches of foreign banks. In addi-tion, 19 foreign banks were established under the off-shore Bangkok International Banking Facilities, whichlent to residents in foreign currency. Total assets of thesystem amounted to the equivalent of 190 percent ofGDP. Commercial banks alone accounted for 64 per-cent of total assets, while finance companies accountedfor 20 percent of total assets. State-owned specializedbanks accounted for a further 10 percent.

In all countries, except Indonesia, banks were mostlyprivate, with many of the larger private banks publiclylisted. The degree of ownership concentration differedacross countries: in Korea, a single ownership limit of 4percent meant that banks were owned by diversegroups of individuals, while in Thailand, despite a sim-ilar rule, several of the large banks were owned or con-trolled by family groups. Similarly, in Indonesia,Malaysia, and the Philippines, banks were owned orcontrolled by corporate conglomerates.

At the same time, a significant part of the bankingsector was owned by the public sector, especially inIndonesia where public sector banks accounted forover 40 percent of total assets of the financial system.In Korea and Thailand, state-owned institutionsrepresented about 15 percent of total assets. InMalaysia one large commercial bank was govern-ment-owned.

There were also foreign banks (branches or sub-sidiaries) with substantial stakes (5–20 percent of totalbanking system assets) in the domestic financialsystem in all five countries, although the degree offinancial openness in each country differed, withKorea, Malaysia, and Thailand being the most restric-tive and the Philippines the least. In Indonesia, foreignbanks were allowed to own up to 85 percent of a jointventure.

Box 3. Structure of the Financial System at the End of 1996

1Assets include trust accounts, which are a significant shareof commercial bank assets (about 40 percent). Trust accountshave grown rapidly and are subject to weaker regulation andfewer restrictions than regular bank accounts.

II VULNERABILITIES

pines were about 150 to 200 percent.14 With the ex-ception of Indonesia, corporate leverage had in-creased significantly over the 1990s.15 The high cor-porate leverage in Korea was largely the outcome ofgovernment policies that emphasized aggressive ex-port-oriented growth and included measures such asdirected credit, subsidized loans, and explicit or im-plicit credit guarantees that biased funding in favorof borrowing rather than equity placements. In Thai-land, corporate borrowing reflected optimisticgrowth projections and a push to gain market share.At the same time, equity markets in the three crisiscountries were undeveloped, since large family-controlled corporations were hesitant to raise fundsthrough equity financing lest their control be diluted.In Malaysia, the stock market was particularlyhighly developed (with a capitalization of over 300percent of GDPat the end of 1996; see Table 2), re-ducing the need for bank financing.

Structural Vulnerabilities

Bank lending practices in the five countries havetraditionally relied on collateral rather than credit as-sessment and cash flow analysis, making banks es-pecially vulnerable to excessive risk taking and de-clines in asset values. For example, during the yearsof high economic growth, credits were increasingly

used to fund investments that turned out to be eco-nomically unsound. When the exchange rate devalu-ation and contraction in demand rapidly erodedcompanies’repayment capacity (see footnote 10),banks and supervisors were suddenly faced withsharply increasing nonperforming loans, loan-lossprovisioning needs, declines in collateral values, anderoding capital bases. Other inadequate lendingpractices including connected lending, high expo-sure to individual clients, and excessive sectoralconcentration of loans, aggravated the problem.

In addition, ineffective market discipline allowedexcessive risk taking. Inadequate accounting anddisclosure practices (see Box 4) and implicit govern-ment guarantees weakened market discipline. A tra-dition of forbearance and “lifeboat” schemes fornonviable institutions instead of firmer correctiveaction or government intervention encouraged ex-cessive risk taking, increased moral hazard, and pre-vented market agents from exerting discipline. As aresult, risk premiums, credit ratings, and analyst re-ports, including reports of international financial in-stitutions, indicate that market participants did notidentify the weaknesses and did not predict crisis.

In the crisis countries, prudential regulation andsupervision had serious deficiencies. These deficien-cies included lax prudential rules, or application ofrules, particularly on connected lending, loan con-centration, cross guarantees, and foreign currencymismatches. A significant problem was the lack ofstrict loan classification criteria and weak rules onloan provisioning and interest suspension. In addi-tion, financial sector regulators and supervisorslacked autonomy, making them susceptible to politi-cal and industry pressure. Supervisors frequentlywaived prescribed limits, a significant problem in

14

The following shortcomings in accounting and dis-closure practices undercut market discipline and fueledthe crisis.

• High corporate leverage was hidden by related-partytransactions and off-balance sheet financing.

• High-level foreign exchange risk exposure by corpo-rations and banks resulting from large, short-termborrowing in foreign currency was not evident.

• Disclosure of loan classification, loan-loss provision-ing, and accrual of interest was weak. Although mostbanks disclosed the accounting policy for loan-lossprovisioning, they did not disclose in the balancesheet the aggregate amount of loans and advances forwhich they had stopped accruing interest.

• In Korea, the practice of cross-guarantees made ithard to assess the solvency of the largest borrowers.

• Consolidation of accounts was generally absent.

• Detailed information on sectoral concentration waslargely absent, even though all countries had largeexposure limits in place.

• Disclosure regarding derivative financial instrumentswas weak.

• Contingent liabilities of the parent of a conglomerate,or of financial institutions, for guaranteeing loans(particularly foreign currency loans) were generallynot reported.

Box 4. Weaknesses in Disclosure Practices in the Asian Crisis Countries

14In comparison, such ratios were about 110 in the UnitedStates, 140 in Germany, and 200 in Japan; see the World Bank(1998).

15For example, between 1991 and 1996, leverage had doubledin Malaysia and Thailand; see the World Bank (1998).


Korea and Thailand.16 The prudential frameworkalso failed to keep up with the developments in thesystem. For example, in some countries prudentialrules on foreign currency exposures failed to limitexcessive foreign open positions or maturity mis-matches. More broadly, building up supervisory ca-pacity was not a priority during the boom years.

Weaknesses in supervision were compounded bythe close links between governments and financialinstitutions. The government’s interference in creditallocation in Indonesia and Korea (through directedcredits) circumvented the need for thorough risk as-sessment by the banks, made the governments co-–responsible for the quality of banks’assets, andprovided an implicit government guarantee onbanks’liabilities. Furthermore, given the govern-ments’ historic role of promoting investmentthrough policy loans and guarantees to corporations,supervisors were constrained in their ability to pe-nalize banks for making bad loans.

15

16In addition, in Korea, commercial, development, and mer-chant banks were regulated and supervised by different agencies,allowing for regulatory arbitrage and making consolidated super-vision difficult.

In all the countries discussed in this paper, urgentmeasures had to be taken to contain the crisis

(Indonesia, Korea, and Thailand) or prevent growingpressures from developing into a full-blown crisis(Malaysia and the Philippines).

In systemic banking crises, major government in-tervention is required even in countries stronglycommitted to market-oriented policies. Such inter-vention is justified by negative externalities associ-ated with widespread bank failures, such as a break-down in the payment system, disruptions to creditflows, and depositor losses. Moreover, financial sec-tor soundness facilitates macroeconomic stabiliza-tion and creates the conditions for the resumption ofgrowth. In all the crisis countries and Malaysia, theauthorities weighed the effects of these externalitiesagainst the potential fiscal costs of intervention andthe moral hazard problems and decided to imple-ment proactive restructuring strategies.

Strategies for dealing with the financial sector cri-sis have sought to incorporate good practices frominternational experience and have the followingcomponents: stabilization of the financial system;changes in the institutional framework to deal effec-tively with the crisis; resolution of nonviable finan-cial institutions; strengthening of viable financial in-stitutions; management of nonperforming assets;and restructuring of the corporate sector (see Box 5for an overview of bank resolution procedures).While the broad strategies were similar across thecrisis countries, each country adapted them to takeinto account national circumstances and preferences(see Appendices I–Vfor a detailed description ofcountry-specific circumstances). These componentsare discussed in detail in Section V.

The initial priorities of such a strategy were to sta-bilize the financial sector and lay out a restructuringstrategy. They were complemented by a macroeco-nomic stabilization plan. Stabilization of the finan-cial sector was accomplished by providing centralbank liquidity support and a blanket guarantee ondepositors and most creditors.17 To stabilize foreign

funding, countries used voluntary debt restructuringwhere feasible (notably Korea), and capital controlson outflows (Malaysia). To cut the flow of centralbank liquidity support, prevent further losses, anddemonstrate their commitment to implement neces-sary reforms, authorities closed institutions judgedto be insolvent or nonviable. These measures werethe first elements of broader restructuring plans.

Macroeconomic Policies

A credible macroeconomic stabilization programwas essential to restore depositor and creditor confi-dence.18 After the initial shocks—that is, with-drawals of foreign funds and exchange rate depreci-ation leading to further withdrawals of capital—allcountries sought to implement macroeconomic sta-bilization policies. Monetary policy was used todampen overshooting of nominal exchange rates andavert depreciation–inflation spirals. Following theinitial depreciations, however, uncertainty over thesuccess of stabilization efforts and continued effortsof foreign creditors to cut their exposure in Asia ledto continued capital flight, further exchange rate de-preciation, and higher interest rates, all of which ag-gravated problems in the corporate and financialsectors. In Indonesia, confidence was further under-mined by policy reversals.

A well-designed, comprehensive, and credible fi-nancial sector restructuring strategy was necessaryfor a sustainable macroeconomic stabilization andresumption of high growth. Progress in structuralreforms was also critical for improving domesticand foreign confidence in these economies. Ques-tions arose over which to put first: should the eco-nomic programs have focused exclusively onmacroeconomic policies, leaving the structural re-

III Addressing the Emergency

16

because of uncertainty about the legal status of the measure andabout the government’s ability to honor it. In Indonesia, the gov-ernment initially announced a limited deposit guarantee that soonhad to be replaced by a blanket guarantee.

18See Lane and others (1999) for a more detailed discussion ofmacroeconomic policies in the crisis countries.

17In Korea, the government initially announced a guarantee onforeign debt, but this failed to stem capital outflows, probably

Macroeconomic Policies

forms for a later time, or should structural reformshave been made at a slower pace? Several consider-ations regarding the financial sector argued againstdelay. First, a banking system saddled with largeamounts of nonperforming loans would have main-tained an excessively cautious lending policy,which would have caused an even greater creditslowdown and further delayed the restoration ofnormal credit flows. Second, where banks (andcompanies) were insolvent, allowing them to con-tinue operating without restructuring would have al-lowed market distortions and moral hazards tobuild. Third, bank and corporate restructuring wasnecessary to facilitate the rollover of maturing for-eign loans and new private investment that was cru-cial to ensuring the necessary financing of theeconomies; it would have been difficult for such

flows to resume if domestic banks and corporationswere perceived to remain financially shaky and in-adequately supervised. Fourth, keeping insolventbanks (and companies) in operation could have en-tailed higher fiscal costs and further complicatedmonetary management.

The crises themselves created a demand for struc-tural reforms. There was a widespread perceptiondomestically and abroad that serious structural flawsin banking and corporate practices had been key de-terminants of the crisis. Thus, economic programsthat failed to address those flaws and practiceswould likely have been viewed to be incomplete orhave only a temporary success. Moreover, there wasa momentum and a social pressure for reform. Forinstance in Korea, labor unions demanded that thechaebols—the large diversified industrial group-

17

A variety of resolution procedures have been em-ployed in the Asian crisis countries. This box definesthe terminology used in this paper.

A bank closure is the act whereby a bank is physi-cally closed to the public and, thus, prevented fromdoing business. A closure can be final or temporary (itmay also be partial, involving continued managementof existing assets and liabilities). In a legal final clo-sure of an institution, there are several resolution op-tions: the institution can exit the system either throughliquidation or through a complete or partial transfer ofits assets and liabilities to other institutions, as dis-cussed below. In a temporary closure, the terms sus-pensionand freezemay also be used. The purpose of atemporary closure is to allow time for a more carefulevaluation of the institution’s situation, or to allowowners time to present recapitalization and restructur-ing plans.

Interventionby the authorities in insolvent or nonvi-able institutions refers to the authorities’assuming con-trol of a bank, taking over the powers of managementand shareholders. An intervened bank usually staysopen under the control of the authorities, while its fi-nancial condition is better defined and decisions aremade on an appropriate resolution strategy. Resolutionstrategies include liquidation, nationalization, merg-ers/sales, purchase and assumption operations, and theuse of bridge banks.

Liquidationis the legal process whereby the assets ofan institution are sold, its liabilities are settled to the ex-tent possible, and its license is withdrawn. A bank liqui-dation can be voluntary or forced, within or outsidegeneral bankruptcy procedures, and with or withoutcourt involvement. In a liquidation assets are sold topay off the creditors in the order prescribed by the law.In a systemic crisis with several institutions to be liqui-dated simultaneously and quickly, special proceduresor special institutions may be needed for the liquida-

tion, as existing structures (e.g., the regular court sys-tem) cannot carry out the job in a timely manner.

Nationalizationmeans that the authorities take overan insolvent bank and recapitalize it. It differs from thetraditional use of the term “nationalization,” which de-scribes a government takeover of a solvent privatebank. Governments in the crisis countries distinguishsuch temporarily nationalized banks from other state-owned banks and often seek to divest/privatize the na-tionalized institutions at an early date.

In a merger (or sale) of an institution, all the assetsand liabilities of the firm are transferred to another insti-tution. Mergers can be voluntary or government assisted.A key issue is to avoid situations in which a merger ofweak banks results in a much larger weak bank, or inwhich an initially strong bank is substantially weakened.

In a purchase and assumption (P&A) operation, asolvent bank purchases a portion of the assets of a fail-ing bank, including its customer base and goodwill, andassumes all or part of its liabilities. In a publicly sup-ported P&Aoperation, the government typically willpay the purchasing bank the difference between thevalue of the assets and liabilities. Variations of P&Aoperations could be a purchase of assets, entitling theacquiring bank to return certain assets within a speci-fied time period, or a contractual profit/loss-sharingagreement related to some or all the assets. P&Aopera-tions in the context of bank resolution can involve theliquidation or transfer of bad assets to an asset manage-ment company.

A variation of a P&Aoperation involves the use of atemporary financial institution—a bridge bank—to re-ceive and manage the good assets of one or severalfailed institutions. The bridge bank may be allowed toundertake some banking business, such as providingnew credit and rolling over existing credits. Bad assetswould be liquidated or transferred to an asset manage-ment company.

Box 5. Bank Resolution Procedures: Terminology and Definitions

ings—be reformed not only on economic groundsbut also on equity considerations. In particular, theunion view that the unavoidable social cost of thecrisis be borne also by the owners of the chaebolshelped to mobilize government support for the re-forms. In Thailand, there was also widespread ac-ceptance of a need for changes to the financial andcorporate infrastructure to ensure that excessiverisks and vulnerabilities of the kind that had led tothe crisis would not be repeated. The sharp reductionin demand put pressure on corporations in all thecountries to restructure their business. For sometime, the authorities had identified and preparedmany of the structural reforms that became part ofIMF-supported programs; the crisis brought thepressure for many such reforms to be implemented.For instance, in Korea, a program of financial re-form had been prepared but was only implementedwhen the crisis broke.

It is doubtful that the crisis economies could havebeen stabilized and confidence returned—even tem-porarily—without implementing major structural re-forms. Malaysia, which did not have an open liquid-ity crisis, also found it essential to implementfar-reaching reforms of its bank and corporate sec-tors. Without structural reforms, forbearance regard-ing loss recognition would have allowed inefficientand unsound enterprises and banks to continue oper-ations, leading to growing distortions, discouragingnew private investment, and constituting a majorburden on economic growth. As a result, the fiscalcosts of the restructuring—which already raisemedium-term sustainability concerns—would havebeen even higher. Also, not addressing the keysources of the crisis would have cast a cloud over thesuccess of any program. All these elements suggestthat, had the structural reforms been delayed or veryweak—for example, the continued financing of non-viable institutions—this would have cast strongdoubts on the sustainability of any macroeconomicadjustment.

Liquidity Support

The central banks in all five countries providedliquidity to financial institutions to offset the with-drawal of deposits and credits at some institutions.Many banks were subject to withdrawals both fromdomestic depositors and creditors, as well as exter-nal creditors.Central banks provided liquidity undervarious emergency lending and lender-of-last-resortfacilities. The amounts were especially large in In-donesia and Thailand (Table 3). Most liquidity sup-port was in domestic currency except for Korea,where the Bank of Korea also provided support inforeign currency ($23.3 billion) to commercial

banks. In Korea, Malaysia, and Thailand, supportwas also provided to nonbank financial institutions,such as merchant banks and finance companies.

To preserve monetary control these massiveamounts of liquidity support had to be sterilized.Sterilization enabled central banks to recycle liquid-ity from banks gaining deposits to those losing de-posits and credit lines. Sterilization had to take placeamid underdeveloped money and interbankmarkets.19 Sterilization was largely effective inKorea and Thailand but not in Indonesia where, forseveral months, protracted political and macroeco-nomic uncertainties resulted in continued depositwithdrawals and capital outflows from the system asa whole, making it impossible for the central bank torecycle liquidity. The resultant highly expansionarymonetary policy led to a continued flight from thecurrency and to the collapse of the rupiah. Since July1998, when overall conditions stabilized, monetarypolicy exercised through market-based auctions be-came more effective. In Malaysia, sterilization waspartial, because of concerns about the effect of highinterest rates on economic activity.

Blanket Guarantees

To stabilize banks’funding and prevent bank runs,Indonesia, Korea, Malaysia, and Thailand an-nounced full protection for depositors and creditors.In all four countries this blanket guarantee was intro-duced as soon as the severity of the crisis becameapparent. Such a guarantee entails a firm commit-ment by the government to depositors and mostcreditors of financial institutions that their claimswill be honored.20 A blanket guarantee generallyaims at providing confidence in the banking system;stabilizing the institutions’liability side; buying timewhile the restructuring work is being organized andcarried out; and preserving the integrity of the pay-ment system. Thailand had announced the major ele-ments of the guarantee in July 1997, which was re-confirmed under the IMF-supported program inAugust 1997. Korea established a full guarantee inNovember of that year, before negotiations with theIMF had started. In Indonesia, the blanket guarantee

18

III ADDRESSING THE EMERGENCY

19Low levels of public debt meant that there was a lack of gov-ernment paper—usually a core element of well-developed moneymarkets. In Indonesia, the development of a market for centralbank bills had been stunted by the authorities’failure to allow fullmarket determination of interest rates.

20The guarantees were, as a rule, not applied to shareholdersand holders of subordinated debt. In Indonesia, insider depositswere not covered by the guarantee. In Thailand, directors’and re-lated persons’deposits and/or claims were not covered by theguarantee unless they could prove that these transactions were at“arms length.”

Blanket Guarantees

was established as part of an IMF-supported pro-gram (January 1998), after a limited guarantee hadfailed to stabilize the situation. Although never facedwith a full-blown crisis, the government in Malaysiaintroduced a blanket guarantee in January 1998. ThePhilippine authorities, in contrast, have not seen aneed for a blanket guarantee. The country had awell-established limited deposit insurance schemethat had been tested in the precrisis period. In noneof the countries was any sort of government guaran-tee extended to entities or shareholders in the nonfi-nancial sector.

The modalities of the guarantees differed slightlyfrom country to country. In Thailand, the guaranteewas preceded by the announcement that the opera-tions of 58 finance companies would be suspendedpending acceptable recapitalization proposals andthat depositors and some creditors in those compa-nies would be compensated in full or in part, in linewith the government’s earlier announcement.21 Alldepositors and creditors of remaining finance compa-

nies and commercial banks were fully guaranteed. InIndonesia, delays in recognizing the systemic natureof the crisis slowed the introduction of the blanketguarantee. Thus, the Indonesian government initiallyattempted to control the crisis by extending liquiditysupport to problem banks and instituting a limited de-posit insurance scheme.22 However, such limited de-positor protection was ineffective, and when a largenumber of banks experienced runs, making apparentthe systemic nature of the problem, the governmentannounced a blanket guarantee for all depositors andcreditors. The Korean government announced a fullguarantee on all depositors and most creditors of fi-nancial institutions. Malaysia offered the blanketguarantee on deposits to all commercial banks, fi-nance companies, and merchant banks, including theoverseas branches of domestic banking institutions.In all countries, the guarantees were announced to betemporary and meant to maintain public confidenceduring the period of restructuring.23All the countries’

19

Table 3. Liquidity Support Provided to Financial Institutions, June 1997 to June 1999

Stock of Support Repaid as of (at peak) Form End of April 1999 Notes

Indonesia 170 trillion rupiah in June 1998 Overdrafts. 10 trillion rupiah.1 Stock of liquidity support(17 percent of GDP). increased from 60 to 170

trillion rupiah between November 1997 and June 1998.

Korea 10.2 trillion in won in December Deposits and loans. 9.2 trillion won. Most of the liquidity support 1997 (2 percent of GDP). provided in November and

December 1997.

23.3 billion in U.S. dollars 20 billion U.S. dollars.(5 percent of GDP).

Malaysia 35 billion ringgit in early 1998 Deposits. n.a. Most liquidity support in early (13 percent of GDP). through mid-1998.

Philippines 18.6 billion pesos (0.8 percent Emergency loans 5.6 billion pesos. Provided in late 1997 to of GDP) in May 1998. and overdrafts. mid-1998.

Thailand 1,037 billion baht in early 1999 Loans from the 54 billion baht.2 Most liquidity support provided (22 percent of GDP). Financial Instititutions in mid-1997 through mid-1998.

Development Fund (FIDF) and capital injections.

Source: IMF staff estimates.1Excluding commitments from former shareholders of banks that received liquidity support to make repayments over four years.2The total would be 561 billion baht, if debt-equity conversions were included.

21A cabinet decision in July 1997 had already guaranteed thedeposits in finance companies. There were two phases in the sus-pension of these companies. In the first phase, 16 companies weresuspended and only depositors were covered by the guarantee. Inthe second phase, 42 companies were suspended, and both depos-itors and creditors were covered by the guarantee.

22This partial scheme had been planned by the authorities forsome time.

23No explicit expiration date was announced in Thailand andMalaysia. The Indonesian government extended the guarantee forat least two years, with a six-month notification period before itwould be lifted. In Korea, the guarantee would expire by the endof the year 2000.

central banks announced that they would provide thenecessary liquidity to make it possible to honor theguarantee.

A blanket guarantee must be credible to stop theneed for liquidity support and the run on banks.Credibility can be enhanced by stating the terms ofthe guarantee explicitly, and by confirming the gov-ernment’s commitment by law and making the guar-antee part of a comprehensive restructuring strategyand part of the macroeconomic program. Most coun-tries faced credibility issues initially. In Thailand,where the full guarantee was announced as part ofthe IMF-supported program, markets did not trustthe government’s commitment until the legal statusof the guarantee had been strengthened, and theguarantee had been tested following the interventionof some banks at the beginning of 1998. In Korea,the comprehensive IMF-supported program bol-stered the credibility of the authorities’rehabilitationplan, including the blanket guarantee. In Indonesia, ablanket guarantee was introduced as part of a newbank restructuring and macroeconomic program inJanuary 1998. As a result of these measures, the ex-change rate stabilized and deposit runs subsidedslowly and stopped after the guarantee was testedduring the closure of seven banks in April 1998.24

While a blanket guarantee may be a necessarycondition to stop bank runs by depositors, it is not a

sufficient one. A blanket guarantee—backed by awillingness to provide the necessary liquidity—canrestore market confidence in a bank’s ability to payback deposits and other protected liabilities in localcurrency. However, if people are fleeing the cur-rency (e.g., because of political uncertainties or eco-nomic turmoil), bank runs will continue because ablanket guarantee cannot restore confidence in thecurrency or prevent capital flight. Also, externalcredit lines may not be rolled over despite the guar-antee, even at higher interest rates.

If the authorities wish to impose losses on deposi-tors and creditors of failing financial institutions, theymust do so before the blanket guarantee is extended.In systemic crises, however, drawing clear distinc-tions between categories of institutions in the initialstages may not be possible due to a lack of informa-tion or equity considerations. In Thailand, the author-ities inflicted losses on some creditors of the sus-pended and subsequently closed finance companies.25

The other countries did not take such a measure andcovered depositors and creditors of all financial insti-tutions still operating at the time of the announce-ment. Indonesia even applied the blanket guaranteeretroactively to the 16 banks closed in October 1997.

Countries have introduced a variety of measuresto limit moral hazard problems. These measures in-clude intensifying the supervision of banks; cappingdeposit rates at a maximum premium above the av-

20


24Subsequently, the guarantee protected the banking system inthe weeks following President Suharto’s resignation in May 1998even though there were further runs on the largest private bank,the owners of which were closely associated with the President,and the rupiah depreciated to its lowest level ever.

25Creditors of 16 finance companies suspended in June 1997were treated more harshly than those of 42 companies suspendedin August.

Indonesiaimposed limits on forward sales of foreignexchange contracts by domestic banks to nonresidents(excluding trade and investment related transactions) inAugust 1997.

Malaysiaattempted to minimize the impact of short-term capital flows on the domestic economy by first re-stricting (August 1997) and later (September 1998)eliminating the offshore ringgit market. As such, themeasures eliminated practically all legal channels fortransfer of ringgit abroad; required the repatriation ofringgit held offshore to Malaysia; blocked the repatria-tion of portfolio capital held by nonresidents inMalaysia for a 12-month period; and imposed tight lim-its on transfers of capital abroad by residents. In Febru-ary 1999, the 12-month holding period rule was re-placed with a graduated system of exit levy onrepatriation of portfolio investments, with the rate ofthe levy decreasing with the duration of investment.

In July 1997, the Philippinesbegan to require priorapproval for the sale of nondeliverable forwards tononresidents and lowered the limit on residents’for-eign currency purchases from banks for nontrade pur-poses. The latter limit was further reduced in April1998.

As soon as the pressure on the exchange rate startedto build up (May–June 1997), Thailand took a seriesof measures to limit baht lending to nonresidentsthrough transactions that could facilitate a buildup ofbaht positions in the offshore market. Genuine under-lying business related to current international transac-tions, FDI flows, and various portfolio investmentswere exempt. These measures in reality led to the cre-ation of a two-tier exchange market with separate ex-change rates for investors who buy baht in domesticand overseas markets (the spreads between both rateswere narrow).

Box 6. Emergency Capital Control Measures

Capital Controls and Debt Rescheduling

erage levels being offered by the “best” banks toprevent weak banks from bidding too aggressivelyfor deposits (in Indonesia and Thailand); coveringthe principal of the deposit only, above a specificthreshold amount (and not the interest); explicitlyannouncing that the blanket guarantee was a tempo-rary measure (Indonesia, Korea, and Thailand); re-quiring institutions to contribute a guarantee fee (In-donesia, Korea, and Thailand);26 and, in case ofinsolvent banks, completely writing down currentowners’shares and removing existing management.

Blanket guarantees entail a very large contingentliability for the government. Initially, the guaranteeis mainly a confidence booster, but by giving a blan-ket guarantee, the government acquires a sizablecontingent liability against assets of uncertainvalue—which most often will be insufficient to payfor the contingent liability that the government willbe called to honor. Even though the blanket guaran-tees entail large costs, these may well be lower thanthe potential economic and social cost of a collapseof the banking system. But since the blanket guaran-tee protects not just small depositors, it may entail aregressive wealth distribution effect because taxpay-

ers’ funds are also used to protect large depositorsand creditors, including external creditors. All thesefactors suggest that in each situation the costs andbenefits of the blanket guarantee have to be weighedcarefully. In case of a systemic crisis, however, ablanket guarantee will be necessary, provided thatthe government can make such a system credible andthat the financial sector is deemed to be sufficientlylarge and of major importance to the economy.

Capital Controls and Debt Rescheduling

To stabilize foreign funding, specific measureshad to be taken. The five countries followed differentpaths to stabilize and reverse the capital outflows.Korea continued to keep its capital account open andrenegotiated the country’s short-term foreign debt. Inresponse to declining rollover rates of short-term for-eign debt, Korea reached an agreement with foreignbanks in January 1998 to reschedule some $22 bil-lion in interbank deposits and short-term loans due in1998. This marked the beginning of the stabilizationof capital flows and of the rapid reduction in centralbank liquidity support. Indonesia, the Philippines,and Thailand imposed temporary capital controlsmeasures to fight currency speculation (Box 6).These controls were lifted in Indonesia and Thailand

21

26In Korea, there is no special contribution to the blanket guar-antee as such, but financial institutions must contribute to theinsurance fund administered by the Korean Deposit InsuranceCompany.

During a systemic crisis, deciding when to closebanks, and which banks, is not easy. The benefits anddrawbacks of immediate closure are outlined below.

Advantages

• Reduces bank losses and minimizes cost to deposi-tors, creditors, or provider of guarantee.

• Ceases central bank liquidity support.

• Allows distribution of losses to shareholders, holdersof subordinated debt, and other creditors.

• Eliminates moral hazard and adverse selection prob-lems associated with insolvent institutions remainingin operation.

• Removes excess capacity from the financial sector.

• Helps to restore confidence in other banks and ingovernment strategy, if done quickly and effectively.

Disadvantages

• Restricts access to deposits and disrupts paymentsystem.

• Severs relationship between lender and borrower,causing a credit contraction.

• Fuels contagion and runs on other banks, if notexecuted properly (for example, without blanketguarantee).

• Leads to further loss of value of bank assets.

• May entail a base money expansion when depositsare paid out.

For closures to be successful

• Must be accompanied by a credible blanketguarantee.

• Owners and subordinated debt-holders must absorblosses.

• The right set of institutions must be included; mar-kets should be reassured that all nonviable institu-tions are dealt with.

• Must include measures to protect payment systemand minimize disruptions to the credit market.

• Must be accompanied by clear and consistent publicannouncements.

• Must be part of a broader strategy to make futurebanking system more efficient.

Box 7. Considerations Regarding the Immediate Closure of Banks in a Systemic Crisis


22

At the outset of the crisis (October 1997), 16 banks(3 percent of assets of banking system) were closed.These banks had been deeply insolvent for severalmonths and had been subject to fraud. Depositor pro-tection was limited to the equivalent of $2,000 (thiscovered over 90 percent of depositors, but only 20 per-cent of the deposit base).

ProblemsThe closure of these 16 banks, unlike in Korea and

Thailand, did not assist in stemming the general loss ofconfidence in the government and the banking sector.The main reasons for this failure appear to be:

• Some politically well-connected banks known to beinsolvent were kept open (despite the recommenda-tions of IMF staff).

• The announcement of the bank closures suggestedthat more banks would be closed later. This news, inconjunction with the lack of a full guarantee on de-positors, led to a flight to safety because depositorsexpected to incur further losses.

• One politically well-connected closed bank was al-lowed to reopen under a new name within weeks,showing ineffectiveness in pursuing the resolutionprocess.

• Failure on part of the government to implement keyreform measures in its IMF-supported program.

• After an initial fall-off of bank runs, depositorsstarted to withdraw funds in sizable amounts frommany banks, with a tendency to redeposit in statebanks.

Box 8. Indonesia: Closure of 16 Banks

–200

–150

–100

–50

0

50

100

150

–25

–20

–15

–10

–5

0

5

10

15MalaysiaIndonesia

Korea

–30

–20

–10

0

10

20

30

Thailand

1997 1998 1999 1997 1998 1999

1997 1998 1999 1997 1998 1999

Blanket guaranteeannouncement

Blanketguarantee

announcement

Closure of7 banks

Rescheduling offoreign debt

Blanketguarantee

announcement

Closure of16 banks,

partial DIS

Intervention in banks

Blanketguaranteereconfirmedby law

Suspension of 42 financecompanies and firstannouncement of blanketguarantee

Suspension of 16 finance companies

–50

0

50

100

150

200

Figure 5. Selected Asian Countries: Central Bank Liquidity Support(Monthly change, as a percentage of reserve money in the preceding month)

Source: IMF, World Economic Outlook.

Immediate Closing of Financial Institutions

once an IMF-supported program was in place.Malaysia introduced more comprehensive controlsover capital flows. Indonesia, Korea, and Thailandall liberalized foreign ownership rules during the cri-sis to attract additional foreign capital to the financialand corporate sectors.

Immediate Closing of Financial Institutions

Closing of insolvent or nearly insolvent financialinstitutions in the three crisis countries was neededto stem accumulating losses and rapidly growing li-quidity support and to give markets a signal thatthere was a break from the past practice of extensiveforbearance (Box 7, see page 21).In Thailand, 58 fi-nance institutions had their operations suspended—56 of which were later closed (for liquidation) afterfailing to present acceptable recapitalization plans.The Korean government initially suspended 14 mer-chant banks, 10 of which were later liquidated. Sub-sequently, seven more merchant banks would beclosed. In Indonesia, the closure of 16 small, deeplyinsolvent private banks was soon followed by inten-sified bank runs. The partial nature of guaranteesprovided to depositors, the perception that otherweak institutions remained in the system, a loss ofconfidence in the government’s overall economicmanagement, and a flight from the currency fueledthe runs. Indonesia’s experience, contrasted withthat of Korea and Thailand, underscores the fact thatbank closures are only successful if all clearly non-viable institutions are closed; the action is part of acomprehensive and credible restructuring strategy;appropriate macroeconomic policies are in place;and the process is clearly and credibly explained tothe public (Box8).

The selection of nonviable institutions to beclosed relied largely on liquidity indicators, such asborrowing from the central bank. This was necessarygiven the typical lack of meaningful information onbank solvency and the realization that insolventbanks can operate as long as they remain liquid. Theliquidity triggers typically included the size of cen-tral bank credit as a multiple of bank capital.27 Onlylater, as more information became available eitherthrough special audits or the supervisory process,could solvency indicators be used as criteria forchoosing nonviable institutions (see Section V). InKorea, solvency data were available from the begin-ning and insolvency was the criterion for the suspen-sion of merchant banks and for corrective action vis-à-vis the two most distressed commercial banks.

The combined emergency measures reduced theneed for central bank liquidity support (see Fig-ure5). In Korea, these measures, in conjunction withthe macrostabilization plan, and the announcementof a restructuring strategy, had an immediate effecton the demand for central bank liquidity support. Asimilar effect can be observed from the announce-ment of the blanket guarantee in Malaysia as part ofa restructuring plan. In Thailand, the demand forcentral bank credit from finance companies abatedin the aftermath of the suspensions, although severalsmall- and medium-sized banks required supportuntil they were intervened in early 1998. As ex-plained earlier, the impact in Indonesia was differ-ent; even after the introduction of the blanket guar-antee, banks’demand for central bank liquidity didnot subside until after the closing of another sevenbanks in April 1998.

23

27For example, four times bank capital was used in Indonesiaand three times in Thailand for the suspension of the 58 financecompanies.

The crisis had profound effects on the overallmonetary environment and on policymakers’

ability to use financial policies to steer the economy.This section examines the difficulties encounteredand discusses the possible interaction between mea-sures to reestablish monetary control and the ob-served decline in credit to the economy in the crisiscountries.

Monetary Management

Monetary management is particularly challengingduring a banking crisis because the relationships be-tween money and intermediate and final targets ofmonetary policy tend to become unstable.28 Bankingcrises can affect the short-run stability of money de-mand, the money multiplier, velocity, the transmis-sion mechanism, and various signa l variables formonetary policy. In the first instance, this occurs be-cause of changes in the composition of money andcredit aggregates.29 Table 4 shows how the variabil-ity of a set of monetary aggregates increased duringthe period. In addition, the segmentation of the inter-bank market in some countries complicated thechoice of interest rate for the central bank to target.(Figure 6 illustrates how interbank rates divergedsignificantly across groups of banks in Indonesia inearly 1998.) In this context, monetary policy focusedon the exchange rate, short-term interest rates andthe level of international reserves (see Section VIIfor a discussion of the implications of the shift in re-lationships for IMF programs).

Credit Crunch

Whether there was a credit crunch in the Asiancrisis countries has been a matter of debate.30 Acredit crunch has been traditionally defined as an ex-cess demand for credit under prevailing interestrates, or a situation where credit is rationed throughnon-price mechanisms. Frequently, however, theterm has been used more loosely to describe a fall inreal credit to the private sector. In all countries, thegrowth rate of real credit has indeed declinedsharply since late 1997, which has been interpretedas a credit crunch (Figure 7). However, care isneeded when measuring and interpreting credit de-velopments in a crisis situation. Measurement ofcredit developments is generally blurred by suchfactors as the treatment of loans that are written offor transferred to an asset management company, therollover of credits, and the effects of bank closuresand valuation changes on the data.31

When an economy is hit by a negative shock, it isoften difficult to determine whether a decline in thegrowth of credit is the result of a shift in demand orin the supply of credit. In Asia, both demand andsupply were affected. On the one hand, demand forcredit declined as consumption and investment weresharply reduced because of uncertainty, overcapac-ity, weakening economic conditions, and the nega-tive wealth effect arising from a fall in asset prices.

IV Monetary and Credit Policies

24

28Garcia-Herrero (1997) discusses the monetary impact of abanking crisis.

29For example, in Korea, credit aggregates would have shown adecline in credit over 1998 unless adjustments were made to takeinto account the large amounts of nonperforming loans being soldto the public asset management company. Monetary aggregatesshould also be adjusted for the large exchange rate valuationchanges.

30For a discussion, see Lane and others (1999), in particularBox 6.5, which, based on several empirical studies, concludesthat the findings are mixed. Empirical work is presented in Ghoshand Ghosh (1999), who found evidence for a credit crunch in In-donesia (in late 1997), and in Korea and Thailand (from late 1997to early 1998), although in the latter two countries, credit demandfell so sharply that supply was not a binding constraint. Dollarand Hallward-Driemeier (1998) found that in Thailand, from late1997 to early 1998, the lack of access to credit was regarded asthe least important factor by manufacturing firms as reasons forslowing down output; Ding, Domaç, and Ferri (1998) found evi-dence for a widespread credit crunch in Thailand, particularly inthe first few months of the crisis.

31For a discussion of these and related issues, as well as en-hancements to the monetary survey to address these weaknesses,see Frécaut and Sidgwick (1998).

On the other hand, borrowers lost creditworthiness,which made banks reluctant to lend, even at higherinterest rates. A self-reinforcing dynamic may de-velop where negative economic shocks may lead toa decline in the demand for credit. Such a situationwill also affect the financial system—leading to adecline in the supply of credit, which, in turn, willaggravate the distress in the real sector, furtherweakening the demand for credit. The followingparagraphs discuss the supply-side factors in moredetail. A full analysis of the demand-side factors isbeyond the scope of this paper.

Even though, in aggregate, deposits did not fall,many institutions had liquidity problems because ofa shift of deposits to higher-quality institutions. Inaddition, the drying up of foreign credit lines forcedbanks to preserve liquidity by recovering assets asquickly as possible and slowing new lending, thusreducing growth in the supply of credit. For exam-ple, foreign claims on banks in Indonesia declinedby about 43 percent between the end of December1997 and the end of June 1998; the decline wasmuch slower (21 percent) during the second half of1998. Corresponding declines for Korea are 27 per-cent and 15 percent for the same periods, and 31 per-cent and 27 percent for Thailand.

25

Credit Crunch

Table 4. Standard Deviation of Selected Monetary Indicators

June 1996 June 1997to June 1997 to June 1998

IndonesiaDemand deposits/M2 0.53 0.95 Time, savings, foreign currency deposits/M2 0.55 0.96 Money multiplier 0.14 0.27

KoreaDemand deposits/M3 0.19 0.45 Time, savings, foreign currency deposits/M3 0.44 0.72 Money multiplier 0.27 0.55

MalaysiaDemand deposits/M2 0.89 1.77 Time, savings, foreign currency deposits/M2 0.91 3.06 Money multiplier 0.09 0.36

PhilippinesDemand deposits/M3 0.50 0.64 Time, savings, foreign currency deposits/M3 0.02 0.02 Money multiplier 0.16 0.21

ThailandDemand deposits/M2 0.30 0.28 Time, savings, foreign currency deposits/M2 0.39 0.51 Money multiplier 0.34 0.58

Sources: IMF, International Financial Statistics; and national authorities.

0

20

40

60

80

100

120

140

2nd 4th3rd 5th 6th 9th 10th 11th 12th

February 1998

Private exchange bank (nonforeign)

Private exchange bank (foreign)

Foreign banks Joint venture banks

State banks

Source: Bank of Indonesia.Note: The market was closed on February 7 and 8.

Figure 6. Indonesia: Daily Interbank MoneyMarket Rates by Type of Bank(In percent per year)

IV MONETARY AND CREDIT POLICIES

The need to increase loan-loss provisions andmaintain capital adequacy affected banks’ability tolend. Stricter capital adequacy requirements and/orprovisioning rules are likely to have further reducedbanks’willingness to lend. Instead of lending bankswould increase their holdings of more liquid andsafer assets, which carry lower weights in the com-

putation of capital adequacy requirements, thus re-ducing the supply of credit.32 To mitigate these ef-

26

32Growth in security holdings other than those acquired as partof recapitalization exceeded the growth in loans for Korea,Malaysia, and Thailand during 1998.

199619951994 1997 1998 199619951994 1997 1998

199619951994 1997 1998 199619951994 1997 1998

199619951994 1997 1998

–40

–20

0

20

40

60Indonesia

–5

0

5

10

15

20

25

30Malaysia

–5

0

5

10

15

20

25

30Korea

–20

–10

0

10

20

30

40

50Philippines

–10–5

0

5

10

15

20

25

30Thailand

Figure 7. Growth Rate of Real Credit to the Private Sector(In percent)

Source: IMF, International Financial Statistics.Note: The spikes in the growth rates between December 1997 and June 1998 in Indonesia, Korea, and Thailand

are due to the effect of the exchange rate depreciation on credit denominated in foreign currency.

Credit Crunch

fects, in each country banks were given time tophase in tighter prudential standards (see Section V).Banks also became extremely risk averse in a situa-tion where the creditworthiness of potential borrow-ers rapidly deteriorated.

The closure of financial institutions may also haveaffected the availability of credit. Customers ofclosed institutions had difficulty building up a creditrelationship with other financial intermediaries in themiddle of the crisis. For instance, in Korea and Thai-land, groups of borrowers served by merchant banksand finance companies lost access to credit as a re-sult of closures. In addition, liquidation processescan have varying effects on credit availability, in thatthey can accelerate repayment (as in Korea), or re-quire customers to repay loans on maturity thatwould otherwise have been rolled over, or, on thecontrary, remove pressure from repayment or servic-ing the debt for some or all customers. To reduce thenegative effects on credit, regulators at first closedonly those institutions that were most deeply insol-vent, applying other resolution procedures thatwould allow customers to continue their credit rela-tionship (such as nationalization, intervention, merg-ers, purchase and assumption operations, or bridgebanks; see Box 5) to remaining institutions.

All countries used a variety of measures in an ef-fort to alleviate the credit slowdown in theireconomies. These include direct measures, such asspecial credit facilities for small- and medium-sizedenterprises, credit guarantees, and mandated credittargets, as well as indirect ones, such as “moral sua-sion” on banks to lend or keep interest rates or inter-

est margins low and gradualism in the applicationof prudential rules and public resources to helpbanks meet their capital adequacy requirements(Table 5).33

Among the measures taken to overcome thebanks’unwillingness to lend was the attempt bysome governments to lower the risk of lending tocertain categories of borrowers by taking over partof the credit risk. Such measures, including guaran-tees on export or import credits or on lending tosmall- and medium-sized companies, may be usefulbut have to be designed so that guarantees do notfully eliminate the banks’credit risk, or relieve themfrom performing a proper credit assessment, even ifthis meansthat guarantee schemes are not fully used.Thus, in Indonesia and Thailand, the partial natureof the export guarantee schemes led banks to morecarefully evaluate credit risks, but it has resulted inundersubscription to these schemes because banksare unwilling to increase exposures to corporationsthat they judge to be insolvent. To address the sameconcern, the Korean authorities intend to phase in re-ductions in the coverage of the credit guarantees of-fered by public guarantee funds, which in most caseshad been 100 percent of the loan.

Measures to alleviate the credit crunch that coercebanks to lend, or that result in the compression orelimination of positive interest margins, can onlyfurther damage already weak banks. Such measures

27

33Some measures that existed before the crisis, such as credit orguarantee facilities for exporters, were augmented.

Table 5. Measures to Alleviate the Credit Slowdown

Measures

Indonesia Credit facility to small- and medium-sized enterprises; allow negative interest ratespreads; Indonesian Bank Restructuring Agency (IBRA) takeover of certain insolventbanks; export credit guarantee scheme; recapitalization assistance.

Korea Discount facility at central bank for loans to small- and medium-sized enterprises;credit guarantee scheme; moral suasion to lend to small- and medium-sizedenterprises; bridge banks and purchase and assumption operations for bankclosures; recapitalization assistance; purchases of nonperforming loans.

Malaysia Moral suasion on banks to lend; lower interest rates (or interest margins); lowerreserve and liquid asset requirements; mandated targets on lending to privatesector; purchases of nonperforming loans.

Philippines Suspended the general provisioning requirement for loans in excess of outstandingstock at the end of March 1999; lower reserve requirements.

Thailand Special credit facility for small- and medium-sized enterprises and exporters; moralsuasion on lending rates; recapitalization support.

Source: IMF.

IV MONETARY AND CREDIT POLICIES

create perverse incentives in that they make banksmore unwilling to lend. Negative spreads in Indone-sia made it profitable for clients to borrow and rede-posit their proceeds (such “round-tripping” forcesbanks to ration credit). Low margins will discouragenew lending and compromise bank profits and their

capital base and, therefore, undermine the entire re-structuring process. The soundest way to alleviate aslowdown in credit is through a combination of mea-sures to stabilize the economy and enhance prof-itability and solvency of banks and their corporatecustomers.

28

The three crisis countries and Malaysia imple-mented comprehensive bank restructuring

strategies. This section discusses selected issues re-lated to the design and implementation of thesestrategies. It reviews broad principles and policiesunderlying such strategies and discusses operationalissues related to the restructuring, such as institu-tional arrangements, issues in valuing financial insti-tutions, the speed of recapitalization, methods todeal with troubled institutions, management ofvalue-impaired assets, the cost of the restructuring,institutional constraints, and linkages with corporaterestructuring.

Broad Principles and Policies

A broad-based restructuring strategy shouldachieve the following economic objectives: (1) re-store the viability of the financial system as soon aspossible so that it can efficiently mobilize and allo-cate funds (a core banking system must be in placeto preserve the integrity of payment systems, capturefinancial savings, and ensure essential credit flowsto the economy); (2) throughout the process, providean appropriate incentive structure to ensure effec-tiveness and, as far as possible, avoid moral hazardfor all market participants, including bank ownersand managers, borrowers, depositors and creditors,asset managers, and government agents involved inbank restructuring and supervision; and (3) mini-mize the cost to the government by managing theprocess efficiently and ensuring an appropriate bur-den sharing (by distributing losses to existing share-holders). To achieve these objectives, governmentshad to ensure effective governance of intervenedbanks, application of appropriate resolution proce-dures, maximization of the value of nonperformingassets, and optimal involvement of private investors.While all the crisis countries followed these broadobjectives, strategies varied according to local cir-cumstances, government preferences, and the depthof the crisis.

Systemic bank restructuring requires strong gov-ernment leadership because the restructuring seeks

to preserve an essential economic infrastructure andentails major macroeconomic and wealth distribu-tion effects, even if in essence it is a microeconomicprocess. Key steps include decisions on institutionalarrangements to deal with the crisis; criteria for eval-uating institutions; a strategy to deal with nonviableinstitutions and to restructure the viable ones consis-tent with macroeconomic goals; the extent andmodalities of public sector support for restructuring;the arrangements for loan recovery and workoutsand asset management; arrangements to ensure oper-ational restructuring; and the pace of restructuringand compliance with prudential norms (see Box 9).Experience also indicates that clear information tothe public on the steps to be undertaken is a crucialpart of the strategy; a nontransparent restructuringprocess may fail to restore the public’s confidence inthe government and the financial system.

The strategies adopted by the crisis countries havebeen broadly similar, in that they all have aimed atremoving nonviable institutions and requiring strictcompliance with international best practices for cap-ital adequacy, loan classification, and loan-loss pro-visioning by the end of the restructuring period(Table 6). All countries aimed at maximizing (do-mestic and foreign) private sector involvement in therecapitalization process. In the event, the extent ofprivate sector involvement has depended on coun-try-specific circumstances, such as the depth of thecrisis, the availability of domestic private fundsamidst a deteriorating macroeconomic situation, andthe legal framework for attracting foreign investors.

Malaysia and the Philippines never experienced afull-blown crisis and applied different restructuringstrategies from those in the three crisis countries. InMalaysia, the emergency measures assisted in con-taining pressures on the system and were followedby a package of proposals that focused on recapital-izing banks that were expected to become undercap-italized in the course of 1998; strengthening the fi-nance company sector through consolidation(mergers); establishing a strong institutional frame-work to manage the restructuring; and strengtheningof regulatory and supervisory frameworks. Facedwith the threat of a crisis, the Philippines adopted a

V Bank Restructuring

29

financial sector reform program in early 1998 tostrengthen the ability of the system to withstandshocks. The main ingredients were a streamlining ofthe resolution procedures of troubled banks, encour-agement of mergers, the privatization of the remain-ing government equity stake in the Philippine Na-tional Bank (the second largest bank in the country),now planned for mid-2000, and an enhancement ofthe prudential and supervisory frameworks.34

The cost of restructuring the financial sector is typ-ically high and largely falls on the public sector. Thisreflects a severe lack of equity capital in the bankingsystem and the corporate sector at the outset of a cri-sis. In the crisis countries, seeking efficient ways torestructure objectives at the least fiscal cost was akey concern of the authorities. A poor fiscal situationcould severely constrain the public sector’s capacityto absorb the cost of the restructure. This was not thecase in Indonesia, Korea, and Thailand, which all hadrelatively sound fiscal positions at the onset of thecrisis.35 However, the immense scale of public sup-port needed will most likely require special efforts topreserve medium-term fiscal sustainability.

Institutional Arrangements

The allocation of responsibilities for handling therestructuring was a crucial first step in the strategy.Taking into account not only technical considera-tions but also political circumstances and institu-tional and legal frameworks already in place, gov-ernments put in place a variety of institutionalstructures:

• In Indonesia, no institution was in charge of re-structuring until the Indonesian Bank Restruc-turing Agency (IBRA) was established in Janu-ary 1998 under the auspices of the ministry offinance. Initial problems in providing adequatelegal and regulatory powers to the IBRAde-layed the effective start of bank restructuringand asset management. Bank Indonesia re-mained the principal supervisory authority,though its powers vis-à-vis state-owned banksand IBRAhad not been clearly defined.

• In Korea, responsibility for restructuring wasgiven to the newly established Financial Super-visory Commission. The Financial SupervisoryCommission also coordinated the work of theother agencies involved in addressing the crisis,including the Korean Asset Management Cor-poration (KAMCO), a bridge bank (HanaerumMerchant Bank), and the Korean Deposit Insur-ance Company (KDIC).

• In Malaysia, a well-designed institutional frame-work supported by strong legislation was al-ready in place. The restructuring has been coor-

30

V BANK RESTRUCTURING

The following issues need due attention in develop-ing a strategy for restructuring banks.

• Institutional and legal frameworks for the restruc-turing, including the allocation of qualified humanresources;

• Criteria for discriminating between those institu-tions that are sound and need no public support,those that are viable but require public support,and those that should exit the system;

• Modalities to assess the financial condition of in-stitutions (deciding on who will do the valuation,and on the valuation rules to be applied, includingloan classification, loan-loss provisioning, and col-lateral valuation);

• Methods for dealing with troubled institutions(liquidation, mergers, nationalization, use ofbridge banks, or purchase and assumptionoperations);

• Treatment of existing and new shareholders;

• Role of government and private (domestic and for-eign) sectors in contributing equity and subordi-nated debt;

• Financing arrangements, including target level ofrecapitalization, types of instruments, terms andconditions for the government’s support of restruc-turing (guided by the principle of minimizing thegovernment’s contribution);

• Arrangements for loan recovery and workouts andmanagement of problem assets;

• Appropriate linkages with corporate restructuring;

• Operational restructuring of banks;

• Timeframe for the different steps in bank restructuring;

• Information campaign and transparency on the re-structuring strategy to ensure credibility and publicconfidence;

• Exit strategy from government ownership ofbanks; and

• Exit strategy from blanket guarantee.

Box 9. Principal Issues in Devising a Bank Restructuring Strategy

34These policies were components of the Stand-By Arrange-ment with the IMF, approved in March 1998, and of the BankingSector Reform Loan with the World Bank, approved in December1998.

35However, the perceived sound fiscal positions concealed thecosts of directed credits, liquidity support to banks, and otherrapidly increasing government contingency liabilities.

Issues in Valuing Bank Assets

dinated by an overarching steering committeechaired by Bank Negara Malaysia, which is alsothe supervisory authority, and includes managersof the three other agencies involved, that is,Danaharta (the asset management company),Danamodal (the bank recapitalization company),and the Corporate Debt Restructuring Commit-tee (the corporate restructuring agency).

• In Thailand, no new agency was set up with spe-cific responsibilities for bank restructuring. TheFinancial Institutions Development Fund, alegal entity within the Bank of Thailand, whichis also the supervisory authority, has been incharge of managing liquidity and solvency sup-port to intervened banks. However, most deci-sion making has been left with the Ministry ofFinance. The Financial Institutions Develop-ment Fund has been hampered by a lack of clearlegal powers. The Financial Sector Restructur-ing Agency (FRA) was set up to assess the via-bility of the 58 suspended finance companies

and to liquidate the assets of the 56 companiesthat were closed. A public asset managementcompany was established to purchase residualassets from FRA. Moreover, the Corporate DebtRestructuring Advisory Committee was set upto facilitate corporate debt restructuring.

• In the Philippines, no new institutional arrange-ments were introduced. The central bank’s rolein bank restructuring has been based on its roleas regulator and supervisor. The Philippines De-posit Insurance Corporation (established in1963) has continued to be involved with the res-olution of insolvent banks; problems of weakbut solvent banks have been addressed by en-couraging mergers.

Issues in Valuing Bank Assets

Realistic valuation of banks’assets is an importantfactor in establishing the viability of individual

31

Table 6. Summary of Measures to Address the Financial Sector Turmoil

Measure Indonesia Korea Malaysia Philippines Thailand

Emergency measuresLiquidity support Yes Yes Yes Yes YesIntroduction of a blanket guarantee Yes Yes Yes No Yes

Institutional measuresEstablishment of an overarching restructuring authority Yes Yes Yes1 No NoEstablishment of a separate bank restructuring authority Yes No Yes No NoEstablishment of a centralized asset management corporation Yes Yes2 Yes No No3

Adoption of a special corporate debt restructuring framework Yes Yes Yes No YesOperational autonomy of restructuring agencies Limited Yes Yes n.a. n.a.

Restructuring measuresIntervention in financial institutions that were weak or insolvent. Yes Yes Yes Yes Yes

This would include:Mergers of weak institutions Yes4 Yes Yes Yes Yes4

Closure of insolvent institutions Yes Yes No Yes YesUse of public funds to purchase nonperforming assets Yes Yes Yes No NoUse of public funds to recapitalize institutions, including: Yes Yes Yes No Yes

State intervention in banks Yes Yes Yes No YesElimination or dilution of current shareholder stakes of

insolvent banks Yes Yes Yes Yes YesNew direct foreign investment Yes Yes Limited5 Yes Yes

Other measuresMeasures to encourage corporate restructuring Yes Yes Yes Yes YesSteps to improve prudential supervision and regulation Yes Yes Yes Yes Yes

Source: IMF.1Steering committee chaired by the central bank.2The powers and resources of a preexisting asset management company were substantially increased.3The Financial Sector Restructuring Agency (FRA) was established to liquidate 56 closed finance companies, and the asset management company to

deal with residual FRA assets.4Between government-owned intervened institutions.5Foreign banks are allowed to purchase up to a 30 percent stake.

banks, but it is difficult during a crisis. In fact, inthese circumstances there is no precise method forvaluing nonperforming loans.36 There are no marketprices for nonperforming loans. Valuation based onappropriately discounted present values becomesless reliable as estimates of cash flows, interest rates,and underlying business conditions become volatile.The valuation can be particularly difficult when theviability and repayment capacity of borrowers is indoubt. Also, it is hard to value collateral, not onlybecause of uncertain prices and a limited market, butbecause of uncertainty as to whether, and when, thebank can seize the collateral.

Dif fering approaches to valuation were used toimprove self-assessments by banks in the three crisiscountries. While banks continued to be responsiblefor valuing their assets and making provisions forlosses, they were also subject to intensified on-siteexaminations by supervisors and assessments by ex-ternal auditors. These on-site examinations and ex-ternal audits generally revealed situations that wereworse than those reported by the banks. In Indonesiaand Korea, these assessments were further supple-mented with audits by internationally recognized ac-counting firms. In Thailand, the authorities ques-tioned the value of additional assessments byinternational auditors of banks meeting all prudentialand regulatory requirements. Each approach has ad-vantages and drawbacks. Self-assessments are oftenbiased due to conflicts of interests; external audits bylocal firms and supervisory evaluations may notcarry sufficient credibility in the market; and foreignassessors may have a limited understanding of a bor-rower’s repayment capacity and other local circum-stances. In addition, international auditors might betoo cautious in their valuations, perhaps to limit theirpotential liability in case they overvalued assets. Re-sorting to international audits, however, seems es-sential for credibility purposes in cases of pervasivegovernment interference or insider lending.

The information collected through either of thesevaluation methods serves as a basic input for the re-structuring authority’s decisions on the viability offinancial institutions.Thus, regardless of the valua-tion methods used, the end result must allow the re-structuring agency to compare banks, based on uni-form and transparent criteria.37 This implies that therestructuring agency or the bank supervisor has tochoose the valuation procedures, including the pos-sible use of external valuation boards. It also implies

that the agency be able to issue regulations on howbanks should assess the value of their assets, but beprepared (and have the power) to overrule valuationsby others where deemed appropriate. This power tooverrule needs to be used judiciously, particularly incases where the assessment has been done by inde-pendent outsiders. Moreover, valuations should besubject to revisions as economic conditions change.In any event, most prospective private investors willundertake their own due diligence valuations prior toany investment in or acquisition of assets or finan-cial institutions.38

Speed of Recapitalization

In all countries, the bank restructuring strategy re-lied on a tightening of rules for loan-loss provisioningand the observance of minimum capital require-ments.39 This gave banks a basis for recognizing theirlosses based on international best practices, identify-ing their capital shortfalls, and putting forward recap-italization plans. The tightening of regulations wasgradual, however.40 On the one hand, markets weredemanding more information about banks’financialconditions and strengthened regulation and supervi-sion. Meeting those demands was viewed as neces-sary for investors to restore the flow of funds to theaffected countries and resume lending and providecapital to domestic financial institutions. On the otherhand, insufficient resources (e.g., capital funds tomeet minimum capital adequacy requirements orlong-term foreign financing to eliminate maturitymismatches) made it impossible for banks to meetstrict prudential standards in the short run. Requiringbanks to meet international standards for capital ade-quacy requirements and loan-loss provisions in a veryshort timeframe would have forced them to shrinktheir balance sheets drastically. This would have fur-ther reduced credit to the private sector and aggra-vated the recession. Thus, a gradual approach wasused. Moreover, it would have been impossible forbanks to effect a reduction in their outstanding loanssufficient to meet the capital adequacy requirement.

Gradualism for achieving compliance with inter-national standards can apply to loan-loss provision-ing or capital adequacy. The former overstates capi-tal adequacy while the latter shows a capital

32


36Nonperforming or value-impaired loans or assets are thosewhose estimated value is below their original book or contractualvalue.

37Valuations should, of course, be based on consistent assump-tions regarding key economic variables and on best practice ac-counting and valuation standards.

38Such due diligence valuations typically take three to sixmonths and are preconditions for investors to buy assets or takestrategic ownership interests.

39These new regulations are discussed in Section VI.40Gradualism, as discussed here, differs from prudential for-

bearance in that the latter refers to the authorities’providing adhoc exemptions and waivers from prudential norms for individualfinancial institutions in a nontransparent way.

Dealing with Troubled Institutions

adequacy requirement below the regulatory mini-mum. Countries have used both approaches. IMFstaff has emphasized that full transparency of thepolicy considerations behind the decisions should beassured to enable investors to make educated deci-sions.41 In Indonesia and Korea, banks have beengiven time to meet their normal capital adequacy re-quirements. The minimum capital adequacy require-ment is currently at 4 percent in Indonesia, but is toincrease to 8 percent by the end of 2001. In Korea,commercial banks were required to meet a capitaladequacy requirement of 6 percent by March 1999and 8 percent by March 2000 (a different schedulewas applied to merchant banks). In Thailand, an 8.5percent capital adequacy requirement for commer-cial banks (8 percent for nonbank financial institu-tions) applies in full while the loan-loss provisioningrequirements are increased each semester until theend of year 2000.42 In Malaysia, valuation and pro-visioning rules were strengthened, but some gradual-ism was allowed with respect to public disclosure ofnonperforming loans. In the Philippines, higher min-imum capital requirements were phased in gradually,aiming at full compliance by the end of 2000.

Dealing with Troubled Institutions

Once nonviable banks were separated from viableones, governments in all crisis countries andMalaysia devised strategies to rehabilitate those in-stitutions deemed viable.To minimize the fiscal costfor the government and to preserve private owner-ship of banks, each government encouraged banks torehabilitate themselves. In cases where market-based solutions were not forthcoming, governmentssought to assist in forging such solutions. In case ofinsolvency, governments intervened. The degree ofgovernment involvement largely related to the de-gree of insolvency of the banks.

The main vehicle for seeking private-sector-basedresolutions was for the respective governments to re-quest recapitalization and rehabilitation plans fromexisting shareholders. In all countries, owners of un-dercapitalized banks were requested to providetimetables to raise the banks’capital adequacy re-quirements to prescribed levels and to show their vi-ability.43 In Korea, the government requested from

banks with capital adequacy requirements below 8percent self-improvement plans to reach that thresh-old, including contributions of new capital from ex-isting or new shareholders. Approval of those planswas a requirement for banks to keep their licenseand for them to receive public sector supportthrough the sale of nonperforming loans to KAMCOor in the form of equity. The precise content of indi-vidual plans varied depending on the circumstancesand the size and significance of the institution.Memoranda of understanding between the banks andsupervisory agencies were used to document the ap-proval of the plans and the conditions attached tothem. The conditions typically included operationalimprovement benchmarks on matters such as costreduction, labor shedding, and rate of return on as-sets. Likewise, the Bank of Thailand requested half-yearly capitalization plans from all undercapitalizedinstitutions, spelling out how they would bring in eq-uity (domestic and foreign) to meet their capital ade-quacy requirements. These plans were agreed uponunder binding memoranda of understanding with theBank of Thailand.

The initial lack of private capital in the three cri-sis countries forced the governments to promoteplans whereby new private capital contributionswould be matched in varying proportions by thegovernment. Under Indonesia’s joint recapitaliza-tion program, for banks with a capital adequacy re-quirement between +4 percent and –25 percent,owners have to submit a business plan demonstrat-ing medium-term viability, in addition to passing afit-and-proper test. Schedules to eliminate excessconnected lending also had to be agreed upon. Own-ers had to provide 20 percent of the capital shortfalland the government the remaining 80 percent.44

Korea followed a case-by-case approach, underwhich the government was prepared to arrange forKAMCO purchases of nonperforming loans, pur-chase subordinated debt, or subscribe new capital,to assist private banks’recapitalization efforts. InThailand, the government will match any amount ofcapital injected by private investors, provided(1) the bank has brought forward and fully imple-mented the end of year 2000 loan classification andprovisioning rules; (2) the new capital (public andprivate) is injected with preferred status; (3) thegovernment and the new investor have the right tochange management; and (4) an acceptable opera-tional restructuring plan has been presented to theauthorities, including procedures for dealing withnonperforming loans and for improving internal

33

41In principle, showing the actual capital adequacy requirementis more transparent and is in line with international accountingstandards.

42However, full up-front application of final provisioning rulesis required in cases where banks seek public funds to match newprivate equity contributions.

43In Korea and Thailand, this was also done for institutions that hadbeen suspended, to give them a last chance to prove their viability.

44In these banks, owners will keep day-to-day control of theirbanks and have first right of refusal to buy back the government’sstake at the end of three years.

control and risk management systems.45 Thisscheme has contributed to restoring confidence inthe Thai banking system, and hence, inducing pri-vate banks to find private investors with or withoutthe public matching funds.

To facilitate foreign participation in the restructur-ing process, governments have liberalized regula-tions on foreign ownership of financial institutions.In addition to bringing in foreign capital, these mea-sures have also aimed at introducing internationalbanking expertise into the domestic financial systemto enhance competition. All countries have allowedforeign investment in existing financial institutions.In Indonesia, two sizeable private banks have re-cently been bought by foreign banks and further pur-chases are expected. In Korea, the banking law hasbeen changed to allow foreigners to acquire a con-trolling interest in domestic banks, including fullownership, and the government is seeking to sell acontrolling interest to foreign investors in two of thelargest commercial banks. In addition, foreign in-vestors have contributed capital to other major com-mercial banks and have started negotiations to investin other segments of the financial sector, such as ininsurance companies. In Malaysia, foreign share-holders are allowed to take a stake in domestic bank-ing institutions to up to 30 percent. In Thailand, for-eign ownership in excess of 49 percent in existingbanks has been allowed to help restructure the sys-tem. Strategic foreign investors have taken a major-ity stake in two small private banks, and foreignownership in some large banks is approaching 50percent. Malaysia has indicated that foreign bankswith operations in the country will be allowed to buyfinance companies. The Philippines has not yet beenfully opened to new foreign ownership of existingbanks, but further liberalization is planned. Recentlydrafted legislation has been submitted to allow 100percent ownership of distressed banks, but this is tobe reduced to 70 percent over 10 years.

When self-rehabilitation was beyond reach, gov-ernments resorted to a variety of bank resolutionmethods to deal with troubled institutions. Suchmethods included interventions, nationalizations,mergers, purchase and assumption operations, andthe use of bridge banks.

A general principle in resolving troubled institu-tions is that existing shareholders should bear lossesuntil their capital has been fully written off. Thisprinciple was generally applied in the crisis coun-

tries, although in some countries shareholders wereleft with nominal stakes to take into account legal re-strictions on a full write-down or to avoid costlylegal challenges by the old shareholders. In Indone-sia, for example, the shareholders in the largest banktaken over in April 1998 were diluted to 1 percent oftotal equity. In Korea, until amended in mid-1998,legislation prevented shareholder stakes from beingwritten down below the minimum capital requiredfor a bank to operate. In Thailand, the shares of own-ers in intervened banks have been written down totoken values.46 Furthermore, in Indonesia, the au-thorities are pursuing former shareholders of failedbanks for personal indemnification for past centralbank liquidity support in those cases where bankshave been in violation of prudential regulations, es-pecially for connected lending.

The contribution of new shareholders is of keyimportance to help strengthen bank finances andgovernance. All countries have revised the rules andregulations governing new shareholders. Existingshareholders are required to meet fit-and-propertests to remain eligible, and rules regarding conflictsof interest for shareholders have been strengthened.A key issue is the maximum size of the equity shareof each individual shareholder: concentration of eq-uity may facilitate governance and capital injections,but concentration may also lead to excessive con-nected lending and large exposure risks. New oramended banking laws in Indonesia and Korea ad-dress this trade-off. In Korea, for instance, the lawlimits the maximum shareholding stake of domesticresidents in commercial banks unless that stake ismatched by a foreigner’s stake.

In the three crisis countries, deep insolvency ofprivate banks led to the nationalization of a signifi-cant part of the private bank sector. In Indonesia,IBRA has acquired control of 12 banks, representing20 percent of the banking sector. The authoritiescontinue to distinguish between the “banks takenover,” however, and the seven state banks existingbefore the crisis, because the aim is to resolve theformer through privatization, mergers, or closureswithin a relatively short period. In Korea, public eq-uity support was very extensive because the limitson single ownership of commercial banks meant thatthere were no significant strategic shareholders thatcould be called upon to inject funds into the banks.Thus, five of the six major corporate lending bankshave ended up with government shareholdings in ex-cess of 90 percent. In Thailand, public equity sup-port has mainly been provided to the institutions that

34


45The new investor has the option to buy back the governmentshares within three years at the government’s initial cost plus car-rying costs. If after full provisioning, the Tier 1 capital adequacyrequirement falls below 2.5 percent, the government will recapi-talize the bank up to this level before matching private funds.

46A legal restriction has prevented the writing down of subordi-nated debt except in cases of legal closure and liquidation.

Dealing with Impaired Assets

were state-owned at the outset of the crisis, to the sixcommercial banks and 12 finance companies thathave been intervened, and to match private equitycontributions in a few private banks.47 All the gov-ernments have expressed their commitment to priva-tize their state-owned banks as soon as feasible.Korea and Thailand have already made someprogress in this direction.

Closures, mergers, purchase and assumption oper-ations, and bridge banks were useful techniques toconsolidate the financial sectors in most countries,and governments adopted them flexibly under thecircumstances. As mentioned before, closures werean important measure in all three crisis countries, asindicated by the proportion of the closed entities inthe sector. Government-assisted mergers were usedin all countries to consolidate the banking system. InKorea and Malaysia, mergers involved private sectorbanks, but in Indonesia and Thailand such opera-tions were limited mostly to the state-owned sector.In Indonesia, four of the seven state banks are in theprocess of being merged into a single bank. In Thai-land, the authorities are merging the intervenedbanks and finance companies into three new banks.The 56 closed finance companies are being liqui-dated by FRAthrough public auctions; the liquida-tion continued through the end of 1999. Mergers,purchase and assumption operations, and bridgebanks have been used in Korea.48 The strategy inMalaysia was different for commercial banks and fi-nance companies. For the commercial banks, a re-capitalization strategy was set up for 14 banks thatwere identified as undercapitalized, or projected tobecome undercapitalized in the course of 1998; inaddition, four banks would be merged into two. Onesuch operation, involving two banks, has been com-pleted. For the finance companies, the governmentinitially aimed at consolidating the 39 companies toless than half that number through mergers. As ofAugust 15, 1999, 15 had been absorbed or merged.49

In the Philippines, private-sector-led mergers wereencouraged through easing of accounting and pru-dential regulations.

In the event, governments used a wide variety ofresolution strategies. In a deep systemic crisis, no

standard solution can be prescribed within the broadoverall restructuring strategy. Governments had todeal with troubled institutions on a case-by-casebasis. The final solution for each institution had totake into account the interest of the parties involved(existing shareholders, potential domestic or foreigninvestors, the government, and creditors) as well asthe legal and regulatory framework, and, often, thepolitical situation. As a result, the outcome for thesector as a whole necessarily varied from country tocountry (Table 7).

Dealing with Impaired Assets

Proper management and disposition of nonper-forming assets is one of the most critical and com-plex aspects of successful and speedy bank restruc-turing. The government’s overarching objectiveshould be to maximize the value of the impaired as-sets in the system, minimize fiscal costs, and preventcredit discipline of borrowers from deteriorating.Various approaches can be adopted to achieve thoseobjectives. Impaired assets may either be held anddealt with by the financial institutions themselves orsold to special companies or agencies created to han-dle bad assets. The likelihood that the borrower willbe able to honor his loan contract should determinewhether the asset should be handled as a loan subjectto collection or as a case for liquidation, includingcollateral. The more likely that the borrower willhonor the loan contract, possibly after renegotiation,the more reason there is to keep it in a bank. If theborrower is bankrupt, or otherwise unlikely to repay,and the bank has to seek recovery of collateral—which often takes the form of real estate or otherphysical assets—a separate institution with specialknowledge in asset resolution techniques most likelyshould undertake the recovery.

The optimal strategy for managing and disposingof impaired assets has many variations, dependingon factors such as the nature of the problem assets,their overall size and distribution, the structure of thebanking system, the legal framework, and availablemanagement capacity in the banks and in the publicsector. There is no single optimal solution but rathera combination of solutions for each country that mayvary over time and for each bank. The strategy willneed to consider the speed of disposition of the as-sets and whether to use a centralized or decentral-ized process and institutional framework (Table 8).

Speed of disposition is determined by the quantity,quality, and type of assets; market demand for suchassets; and whether the assets belong to a bank thathas been closed or to one in operation. While eco-nomic recovery requires some asset sales or liquida-tions to help markets find new price levels, the mar-

35

47The intervened institutions were insolvent and not in a posi-tion to raise capital. Financial Institutions Development Fundtook control, recapitalized them, and is preparing them for repri-vatization.

48In Korea, the bridge bank, a subsidiary of the KDIC, wasgiven a life span of three years and was used only for the good as-sets of closed merchant banks.

49According to the most recently announced merger exercise(July 29, 1999), the domestic commercial banks, finance compa-nies, and merchant banks will be consolidated into six large fi-nancial groups.

kets may be extremely thin and care needs to betaken not to destroy values for the entire bankingsystem through “fire-sale” liquidations. This is ofparticular concern in a systemic crisis when theamount of problem assets typically is very large. Thevalue of impaired assets may be better preservedthrough careful management and gradual sales byspecialized institutions (in this paper all such unitslocated outside banks are referred to as asset man-agement companies). At the same time, it is impor-tant not to “park” severely impaired assets for yearsin asset management companies while waiting for aneconomic upturn. Such an approach may result in ac-crual of carrying costs and ultimately bigger losses.Moreover, poor management of the assets may dete-riorate their value. In the case of failed banks, it isimportant to move the better quality loans to otheroperating institutions as fast as possible to preservevalue. In Korea this was done through bridge banks,while in Thailand most of the assets of the 56 closedfinance companies have been sold to the private sec-tor through public auctions carried out by the FRA.50

Each country considered the advantages and dis-advantages of dealing with impaired assets in a cen-tralized or decentralized asset management companystructure and related ownership issues in its own cir-cumstances (Table 9). Centralized asset managementcompanies, which typically need to be state-owned,have been used in Korea and Malaysia, and more re-cently in Indonesia. Advantages and disadvantagesof state-owned, centralized asset management com-panies are shown in Box 10. A key objective of astate-owned centralized asset management companyis to buy nonperforming loans from banks and thushelp banks clean up their balance sheets as fast aspossible. It is also useful in cases of mergers or banksales when the merging or purchasing party may notwish to get a large amount of nonperforming loansas part of the deal. Thailand has chosen a decentral-ized process, encouraging each commercial bank toestablish its own separate asset management com-pany. However, a public asset management companywas established to purchase residual assets fromFRA. In Indonesia private asset management com-panies to deal with failed banks were ruled out dueto governance concerns.51 The sale of banks’assets

36


Table 7. Mergers, Closures, and State Interventions of Financial Institutions(June 1997 to June 1999)

Mergers Closures State Interventions

Indonesia Four of the seven state commercial 64 commercial banks (18 percent). 12 commercial banks banks to be merged into a single (20 percent).commercial bank (54 percent).

Korea Nine commercial banks and two Five commercial banks, 17 merchant Four commercial banks merchant banks to create four new banks, more than 100 other nonbank (14 percent).1commercial banks (15 percent). financial institutions (15 percent).

Malaysia 15 mergers (6 percent)(finance None. One merchant bank and three companies and commercial bank). finance companies under

central bank control (3 percent).

Philippines Four commercial bank mergers One commercial bank (1 percent).2 None.(2 percent).

Thailand3 Three mergers involving five 56 finance companies (11 percent) Six commercial banks and 12 commercial banks and 12 finance and one commercial bank (2 percent). finance companies (12 percent).companies (16 percent).

Source: IMF.Note: Figures in parentheses refer to percentage of total banking system assets held by the corresponding group of institutions.1Banks with over 90 percent government ownership.The government owns varying amounts of shares in seven other commercial banks.2Closures of a number of rural banks and small thrifts are not included. Such closures are routine operations in the Philippines.3In Thailand, most of the intervened institutions were later merged.Thus, columns one and three include the same institutions.

50So far, more than half of the assets of the 56 finance compa-nies have been sold to the private sector and most of the remain-der will be sold to the public asset management company that wasset up as the residual buyer of all FRAmanaged assets that couldnot be sold in the auctions.

51This was done to prevent associates of some of the failed banksfrom setting up private asset management companies, which couldhave circumvented rules for pricing and prudent governance.


to an asset management company forces immediaterecognition of the value of the loan. This may detersuch sales in cases where banks have been carryingthese loans at inflated values.

Pricing is the most difficult issue for a public assetmanagement company purchasing assets from pri-vate banks. The issue is less severe for a public assetmanagement company buying assets from a state-owned bank or a private asset management companybuying assets from a private bank.52 This is due tothe valuation difficulties for impaired assets dis-cussed earlier. Purchases of a bank’s assets at in-flated values by an asset management companyamount to a back-door recapitalization of the bankand a bailout of the bank’s shareholders.53 The im-possibility of determining an unambiguously fairmarket price for nonperforming loans has so far de-terred the Thai authorities from setting up a public,centralized asset management company and theyhave instead opted for a decentralized approach.Proper transfer pricing is also of key importance forthe incentive structures for both the asset manage-ment company and the banks—there is a need to setup a system that provides the right balance. Exces-sive prices for nonperforming loans may also inducebanks to reduce their recovery efforts, which couldlead to a general deterioration of credit disciplineand loan values throughout the banking system.

A decentralized approach that encourages eachbank to set up its own asset management companyallows arrangements to suit each bank’s conditions(see Table 10). Thailand followed such a route, en-couraging banks to set up their own asset manage-ment companies to which they can transfer assets at

market value. Five private banks are in the processof setting up asset management companies and otherbanks (including state-owned ones) are expected tofollow. Until recently, however, this process hasbeen held up due to capital scarcity.

The final results of the various strategies will onlybe known when the process of recovering impairedassets has been completed. This process will taketime. As noted earlier, it would be unwise to under-take massive sales of assets in the midst of the crisis.Also, practical problems—such as the need to ac-quire proper legal title to collateral and to prepare aninventory of the assets—require time to be solved.Nonetheless, sales of impaired assets have begun inKorea, Malaysia, and Thailand.


Estimating the cost of financial restructuring isone of the more challenging issues. There are costsboth in the private and public sectors to cover lossesand contribute new capital. The private sector out-lays will not be considered here. The government’sgross costs for the restructuring arise from payingout guaranteed bank liabilities; providing liquiditysupport; assisting in meeting capital adequacy re-quirements; and purchasing nonperforming loans.The net costswill only be known after proceedsfrom (re)privatization of banks and recoveries ofloans accruing to the government have been takeninto account.54 A more complete picture of the costwould also include the indirect effects of the crisisand subsequent reforms. The magnitude of thesecosts and the need for political support for theprocess require transparent accounting rules and dis-closure of information.

37

Table 8. Framework for Managing Impaired Assets

Speed of Disposition Decentralized Centralized

Rapid (sale or liquidation) Direct asset sales by banks; Rapid resolution vehicles Liquidation of a bank. (Thai Financial Sector

Restructuring Agency, or FRA).

Long-term (asset Individual asset management Long-term disposition management) companies (Thailand); bank (asset management).

workout units (all countries).

52Consolidated accounting needs to ensure that transfers to pri-vate asset management companies are not used as an artificialway of cleaning up the banks.

53In Korea, the asset management company had to change itscriteria for determining the purchasing price, since the purchaseprices determined initially appeared to be too high; even after theadjustment the asset management company had no difficulty ingetting banks willing to sell assets.

54In Indonesia, a significant offset to the cost of restructuring isalso likely to come from the disposal of assets pledged by share-holders of failed banks in violation of legal lending limits. IBRAalso hopes to secure significant recoveries based on forensic ex-aminations of illegal transactions of some of the failed banks.


Gross costs in the three crisis countries andMalaysia are likely to range from 15 percent to45 percent of GDP(Table 11). Estimating the cost ofrestructuring is an evolving exercise because lossrecognition is still taking place as part of the corporaterestructuring process.55 The cost of restructuring willdepend on several factors, including domestic and ex-ternal macroeconomic conditions, the effectiveness ofcorporate restructuring, and the efficiency of bank re-structuring efforts. As a result, estimates for the costof restructuring vary widely, with government num-bers generally lower than market estimates.56 While

IMF staff has continuously made estimates based ondifferent scenarios to discuss policy options with theauthorities, it has refrained from including any esti-mates in official documents because of their sensitiveand crude nature.

The need for immediate liquidity support at theonset of the crisis meant that the central banks in thecrisis countries were the main providers of funds. InIndonesia and Thailand, formal arrangements to al-locate costs between the central bank and the gov-ernment were weak or did not exist. Meanwhile, inIndonesia, the government has issued 150 trillionrupiah (13 percent of GDP) of indexed bonds to thecentral bank to compensate it for past liquidity sup-port. In Thailand, the government was authorized toissue bonds for 500 billion baht (10 percent of GDP)to cover losses in the Financial Institutions Develop-ment Fund, and the government has announced itsintention to cover additional losses in a similarway.

38

Table 9. Public Asset Management Companies in the Asian Crisis Countries

Amount Purchased by theCentralized/Public Asset End of April 1999 (face Management Companies value as a percent of GDP) Eligible Loans Pricing Policy

Indonesia Became fully operational in 20 Loss loans from state banks, Zero value.April 1998 within Indonesian and private banks eligible Bank Restructuring under the government’s jointAgency (IBRA). recapitalization scheme.

Korea Korean Asset Management 10 All financial institutions. 45 percent of face value for Corporation (KAMCO) was secured loans; 3 percent for reconstituted as an asset unsecured loans.management company in late 1997.1

Malaysia Danaharta was established in 17 2 All financial institutions, Average discount has been mid-1998. including Labuan Subsidiaries 37 percent (excluding one

of Malaysian banks and large loan, it has been 60 development financial percent).institutions.

Philippines No asset management . . . . . . . . .company.

Thailand Financial Sector 17.5 Assets of the closed financed Auctions. In a first round Restructuring Agency (FRA) companies. only private bidders and, if established on October 24, not sold, to the asset 1997, to deal with suspended management companies.finance companies.

An asset management company 43 Finance companies or In the second round, the was set up to act as bidder of intervened banks (so far only asset management companies“last resort” for assets of finance companies). purchased at 20 percent ofclosed finance companies. book value on average.

Source: IMF.1KAMCO existed before the crisis.2End of June 1999.3This 4 percent is included in the 17.5 percent.

55Banks’accounts may continue to show additional losses forsome time even if economic conditions improve because the pos-itive effect of recovery on the loan portfolio may be offset bylosses from existing loans.

56For example, Merrill Lynch estimated recapitalization require-ments for commercial banks—just one component of gross costs—at 42 percent for Indonesia, 10 percent for Korea, 11 percent forMalaysia, and 26 percent for Thailand (Merrill Lynch, 1999).

Institutional Constraints

Bonds issued or guaranteed by the governments ofIndonesia, Korea, and Thailand are the main instru-ments for financing the government’s contribution tothe restructuring costs (Table 12). Market interestrates and regular coupon payments are needed, be-cause, as opposed to zero coupon bonds, they helpbanks’cash flows. Tradable bonds help banks man-age their liquidity, as they can sell the bonds if li-quidity is needed.57 Given the large amount of bondsto be issued, making them tradable also assists thedevelopment of a government bond market and re-duces the cost for the government; thus, the develop-ment of an efficient microstructure for governmentsecurities markets becomes critically important.58

Only Korea and Malaysia used a combination ofcash and bonds to provide capital, although cash in-jections in both cases were so minimal that they didnot interfere with monetary policy. The governmentof Indonesia has recently approved the issuance ofanother 100 trillion rupiah of bonds (9 percent ofGDP) to the banks to finance the first wave of bankrecapitalization. The interest cost of all the bonds isborne by the budget. In Korea, parliament approved

the issuance of 64 trillion won of bonds to financeKAMCO and KDIC (15 percent of GDP).

A full and transparent recording of the cost ofbank restructuring is important.While the initialsupport by the central banks was not very transpar-ent, bringing the outlays into the budget wouldimply more transparency. All of the crisis countrieshave incorporated into their budgets the interest pay-ments on the governments’recapitalization bonds.However, the cost of earlier liquidity support and thecapital cost of government bonds have not yet beenaccounted for in the budgets. Incorporating the totalrestructuring costs into the budget is crucial, notonly to have a clear overview of the total cost, butalso to be in a position to better assess the countries’medium-term fiscal sustainability. This can beachieved through the use of an augmented fiscal bal-ance that would explicitly incorporate all majorquantifiable fiscal costs of bank assistance opera-tions (Box 11). For financial programming purposes,however, the carrying costs of financial sector re-structuring is a more appropriate concept.

Institutional Constraints

Restructuring a banking system following a sys-temic crisis of the magnitude experienced in Asiancountries is a complex process that goes far beyondpurely technical operations. Restructuring strategieshave to take into account local business practices,the availability of human resources, the deficiencies

39

The crisis countries considered the following whendeciding whether to adopt a centralized, public assetmanagement company.

Advantages

• Serves as a vehicle for getting nonperforming loansout of troubled banks, based on uniform valuationcriteria.

• Allows government to attach conditions to purchasesof nonperforming loans in terms of bank restructuring.

• Centralizes scarce human resources (domestic andforeign).

• Centralizes ownership of collateral, thus providing moreleverage over debtors and more effective management.

• Serves as a third party for insider loans (Indonesia).

• Can better force operational restructuring of troubledbanks.

• Can be given special legal powers to expedite loanrecovery and bank restructuring.

Disadvantages

• Management is often weaker than in private struc-tures, reducing the efficiency and effectiveness of itsoperations.

• Such agencies are often subject to political pressure.

• Values of acquired assets erode faster when they areoutside a banking structure.

• Nonperforming loans and collaterals are often“parked” long-term in an asset management com-pany, not liquidated.

• If not actively managed, the existence of a publicasset management company could lead to a generaldeterioration of credit discipline in financialsystem.

• Cost involved in operating an asset managementcompany may be higher than a private arrangement.

• If dealing with private banks, determining transferprices is difficult.

Box 10. Advantages and Disadvantages of a Centralized Public Asset Management Company

57Under special circumstances, governments could decide toissue non-negotiable bonds, for instance to prevent insolventbanks from selling those bonds and investing the proceeds inrisky assets (as in Indonesia).

58For a discussion of the development of a microstructure forgovernment securities, see Dattels (1997). The importance ofsmooth functioning money markets for the development of cen-tral bank open market operations is discussed in Quintyn (1997).


in the legal and judiciary framework, and dependlargely on the degree of political support. In thethree crisis countries, the changeover in the politicalregime had a clear positive impact on the pace of re-structuring. Full recognition by the new govern-ments of the magnitude of the crisis not only in-creased their resolve to implement the restructuringstrategy, but also made it easier to gather broad-based support for the restructuring.59

Bank restructuring requires a large number ofpeople with a wide variety of skills. Countries sel-dom have such resources readily available becausecrises are only sporadic occurrences. Governmentsoften had to rely on outside expertise to help developtheir strategy and carry out specialized tasks. In ad-dition, the number of people needed was consider-able. For instance, approximately 5,000 people (for-eign and local experts, officials from differentagencies, and security forces) were involved in thefinal closing of the 56 finance companies in Thai-land. Similar numbers were involved in such opera-tions in Indonesia.

Deficiencies in national legal and judicial frame-works have been major obstacles to the restructuringprocess in the crisis countries. Key issues include theinitial lack of proper and tested exit policies forbanks (especially insufficient authority to take earlyaction against weak banks and to eliminate share-holders); the lack of legal protection of officials (theabsence of immunity from prosecution and civil

suits of officials in the exercise of their duties is stilla major problem in the Philippines and Thailand);the inadequacy of foreclosure procedures and bank-ruptcy laws and an appropriate judicial infrastruc-ture to deal with the massive corporate debt restruc-turing problem; a legal framework that often favoreddebtors over creditors; and the slowness and inexpe-rience of the courts. Once deficiencies were recog-nized, it took a considerable amount of time beforelaws could be changed, and parliaments sometimesintroduced counterproductive amendments, furtherdelaying the process.

40

Table 10. Pros and Cons of Decentralized Asset Management

Advantages Disadvantages

Within banks Knowledge of the borrower may Lack of skills for restructuring of facilitate debt restructuring. troubled debt, operations of Access to borrower through companies, debt-equity swaps, etc.branch network. Hampers “normal” banking functions

(lending activities), particularly if thenonperforming loan portfolio is large.Less loss recognition up front. Doesnot clean up the bank’s books.

In private asset Specialized skill mix. Focus on Lack of knowledge of the borrower.management restructuring function. Bank may not have sufficient capital to companies Creation of an asset management recognize losses up front.

industry and secondary market fordistressed assets.Loss recognition up-front.Cleans up the bank’s books.

59Public recognition of the size of the crisis was delayed be-cause these economies had for long been viewed as highly suc-cessful, and therefore unlikely to face a major crisis.

Table 11. Authorities’ Estimates for theGross Cost of Financial SectorRestructuring

In Percent of GDP

Indonesia 45Korea 15Malaysia 12Philippines n.a.Thailand n.a.

Source: National authorities.Note: Gross costs do not include recoveries.The cost will de-

pend on a number of factors, including macroeconomic condi-tions, effectiveness of corporate restructuring, and the efficiencyof the bank restructuring effort. Estimates of these costs arebased on different assumptions and methodologies and maytherefore not be comparable across countries.

Linkages to Corporate Sector Restructuring

Linkages to Corporate Sector Restructuring

The severity of the corporate sector crisis in thethree crisis countries has affected bank restructuringmore than in most other bank crises. In most crisiscountries, corporate sector restructuring beganslowly and is lagging behind bank restructuring.This stems from the fact that the loss recognitionprocess took longer, legal frameworks for address-ing the issues were not or only partially in place, ad-ditional skills were needed that were not readilyavailable, and the sector itself is more complex anddiversified than the banking sector. Most important,unlike the banking sector, the private, corporate sec-tor is not under the control of one single regulatoryand supervisory agency. Moreover, corporate debtrestructuring (the part that has a direct bearing onbank restructuring) largely depends on a broaderbusiness restructuring, which is usually a slowprocess. To the extent that corporate restructuringcontinues to lag behind, bank restructuring might bedelayed. Generally, the two processes should pro-ceed as simultaneously as possible, although bankrestructuring should take the lead. This is so, notonly because it is more feasible because banks arefewer and are already subject to an established su-pervisory regime, but also because it is necessary to

have functioning financial institutions as counterpar-ties to facilitate the corporate restructuring. Severalinitiatives have been taken to expedite corporate sec-tor restructuring (Box 12).

41

Table 12. Instruments Used to Recapitalize and Purchase Nonperforming Loans

Instrument Received Description

Indonesia Bonds. Equity. Variable and fixed coupon bonds. Owners underthe private bank scheme may purchase backshares after three years; tradable.

Bonds. Nonperforming loans. Variable coupon bonds; tradable.

Debt to equity conversions. Equity. Index-linked bonds; tradable.

Korea Bonds or cash. Equity or preferred shares. Variable coupon bonds issued by KAMCO andKDIC and guaranteed by the government; tradable.

Shares. Equity or preferred shares. Governments share in public enterprises.

Bonds. Nonperforming loans. Fixed (initially) and variable coupon bonds; traded.

Malaysia Bonds or cash. Convertible preferred shares Zero coupon bonds; not easily traded.or subordinated debt.

Bonds. Nonperforming loans. Zero coupon bonds; not easily traded.

Philippines Not applicable.

Thailand Debt to equity conversions. Equity. FIDF loans converted to equity.

Bonds. Equity. Fixed-coupon bonds; tradable.

Bonds. Subordinated debt. Fixed-coupon bonds; nontradable.

Source: IMF.

The concept of augmented fiscal balance wouldexplicitly incorporate the major quantifiable fiscalcosts of bank assistance operations that are not in-cluded in the current definitions of the overall bal-ance.1 The augmented balance is not intended to re-place the overall balance, but to present anadditional measure of the fiscal stance for countrieswhere bank assistance operations are important. Useof the augmented balance would allow a transpar-ent, comprehensive, and reasonably comparablepresentation of government financial assistance forbank restructuring across countries. If not using anaugmented balance framework, complete details ofthe capital cost of bank restructuring operationsshould be recorded separately, regardless of whetherthey are budgetary or quasi-fiscal costs.

Box 11. The Augmented Fiscal Balance

1See Daniel, Davis, and Wolfe (1997).


The banking system provides a key lever forcorporate restructuring, in particular as regardscorporate debt. Tightening and stronger en-forcement of prudential regulations can make amajor contribution. For instance, in Korea, tighterregulations on maximum exposure limits, con-nected lending, liquidity mismatches, and crossguarantees have required corporations to seek other

sources of finance or shrink their balance sheets.Also, Korean banks—following strong governmentleadership—have been instrumental in approvingand monitoring corporate restructuring plans. Inter-national financial institutions, particularly theWorld Bank, have strongly supported this process,including the banks’heavy involvement as keycounterparts.

42

Corporate restructuring is lagging behind bank re-structuring, and has been hampered by a lack of lever-age of most banks vis-à-vis their borrowers. The coun-tries generally lacked frameworks for restructuringfailing borrowers, such as coordinating credit commit-tees under the guidance of a lead bank. Moreover, inmany countries, the courts have tended to be lenientand provided little support for creditors, at least in theinitial stages. Companies have been able to continue tooperate under the court’s protection, allowing interestto be deferred. This has undermined credit discipline.In some countries, measures have been taken to ad-dress this problem: first, through strengthened bank-ruptcy and foreclosure procedures to lessen the bal-ance in favor of the borrowers, and second, byproviding incentives and mechanisms for banks andcorporate clients to restructure before reaching thecourts. Initiatives also have been taken to improvecourt procedures.

The speed at which corporate restructuring takesplace depends on a variety of factors, including thelegal and regulatory framework at the onset of the cri-sis, the vigor of its enforcement, the structure of thecorporate sector, and the nationality of the creditors.Corporate restructuring in Indonesia has so far beenslow because new bankruptcy and foreclosure lawshave not been enforced, in particular by the state-owned institutions. Korea is more advanced than theothers because it adopted new laws fairly rapidly, has acorporate sector concentrated around the chaebols, andhas clearly specified its restructuring objectives andprovided strong leadership. The legal framework isstrongest in Malaysia. Thailand suffered from delays inadopting a new legal framework for bankruptcy anddisclosure.

Countries have given a variety of incentives to banksto address and expedite corporate debt restructuring. The“London Approach,” which provides a frameworkwhereby creditors would jointly approach the debtor andwork out a mutually beneficial (out-of-court) arrange-ment, has so far been used most aggressively for govern-ment-led restructuring in Korea, where a number ofmedium-sized corporate groups have been dealt with inthis way (debt rescheduling with some interest rate re-duction, or issuance of convertible bonds and debt/eq-uity swaps). In Thailand, the Corporate Debt Restructur-ing Advisory Committee (CDRAC) was formed by theBank of Thailand and representatives from debtor andcreditor groups, which agreed upon a framework for cor-porate debt restructuring based on the London Approach(the “Bangkok Approach”). The CDRAC initially tar-geted large debt restructuring cases but has recently ex-panded its coverage to small- and medium-size cases. Ascheme that would combine government support for re-capitalization with corporate debt settlement (Tier 2 op-tions) has also been in place to encourage corporate debtrestructuring. In Indonesia, the government has adopteda four-way classification of delinquent borrowers, basedon their degree of cooperation with the workout processand business prospects. The state institutions and In-donesian Bank Restructuring Agency (IBRA) are adopt-ing a coordinated approach vis-à-vis each major delin-quent borrower, beginning with the largest, broadlyalong the lines of the London Approach (the “JakartaInitiative”). In Malaysia, the asset management com-pany (Danaharta) has been given very extensive powersin dealing with the borrowers of any loans it buys. Thesepowers seem to have had a significant effect on borrow-ers, because banks threatened to sell to Danaharta if theborrower failed to resume servicing the debt.

Box 12. Linkage to Corporate Restructuring

Early in the crisis, all countries took initiatives tostrengthen prudential regulation and supervi-

sion, bank governance, and market discipline. Thenature and direction of these initiatives was similarin all countries: to bring domestic standards closer tointernationally accepted practices. Preparations havealso started for the eventual removal of the blanketguarantees. Initiatives have also been taken to im-prove prudential regulation and supervision of inter-national creditors in their home countries.

Prudential Supervision

In the crisis countries, the autonomy of the super-visory authority has been strengthened. In Indonesia,a new central bank law, passed in May 1999, estab-lished independence for Bank Indonesia. Under thelaw, responsibility for supervision moves from BankIndonesia to a new agency in 2002. In Korea, beforethe crisis, a proposal for a new autonomous supervi-sory authority met with considerable resistance, butafter the crisis broke the legislature swiftly passednew legislation establishing such an autonomousbody. More recently, additional powers for licensingand supervision have been transferred to the new au-thority. In Thailand, supervision has been strength-ened and become more autonomous. The new Bankof Thailand and Financial Institutions Acts are beingdrafted to give the Bank of Thailand greater supervi-sory authority. These laws are expected by early2000.

All countries have made efforts to upgrade theirsupervisory capacity and strengthen the powers ofsupervisors. Supervisory reporting by banks hasbeen much improved, and countries are relying moreon on-site examinations. Moreover, supervisors cannow demand additional loan-loss provisioning frombanks, corrective actions when problems are de-tected, and more support from banks’external audi-tors. The use of memoranda of understanding toenhance the supervisory authorities’ability to moni-tor and enforce compliance with prudential ratiosand performance benchmarks of financial institu-tions has become common. Supervisors now have

more power over entry of banks, supported by newfit-and-proper rules (in all countries, the policygenerally is not to allow new entrants until therestructuring is over). Most of the countries haveintroduced procedures that will remove some earliersupervisory discretion that allowed them to waivecompliance with regulations on a case-by-case basis,for example regarding loan concentration. Tostrengthen supervisors’hands in dealing with prob-lem institutions, mandatory “prompt corrective ac-tion” procedures have been or are being introducedfor situations where capital adequacy falls belowcertain trigger points.

Prudential Regulation

Regulations concerning loan classification, provi-sioning, and income recognition have been broughtcloser to compliance with international best prac-tices (see Tables 13 and 14). Loan classificationrules were strengthened in all countries. The periodoverdue for interest suspension was shortened tothree months in Indonesia and Thailand (this rulehad already been tightened in the Philippines). Allthe countries tightened their specific loan-loss provi-sioning requirements and introduced or tightenedtheir general provisioning requirements. All thecountries, with the exception of the Philippines,have made required loan-loss provisions tax de-ductible. The definitions of capital have been im-proved and, in some countries, the absolute mini-mum levels of capital have been increased. Asdiscussed earlier, banks were given time to complywith these new regulations, according to specifictimetables.

Several other key prudential regulations are beingimproved. These include foreign exchange exposurelimits in all countries; liquidity management rules inIndonesia, Korea, and Malaysia; connected lendingregulations in Indonesia and Korea; single borrowerand group exposure limits in Korea and Malaysia;and cross guarantees within chaebols in Korea. Fit-and-proper rules for owners and managers were in-troduced or strengthened in all countries. In addi-

VI Institutional Reform

43

VI INSTITUTIONAL REFORM

tion, other imprudent banking practices, such asgovernment directed lending, where applicable, arebeing phased out. The speed with which new regula-tions could be phased in had to take into account notonly bank and supervisory agency constraints in in-troducing any new regulation, but also crisis-specificelements, such as the ability of their clients to com-ply with such rules.

Most countries have taken measures to improvetransparency and disclosure, as well as the qualityof data disclosed. Quality of data has been im-proved by new loan classification, provisioning,and income recognition rules and by extensive in-volvement of onsite examiners and internationalauditors and analysts to support banks’recapital-ization efforts. This is expected to enhance gover-nance and market discipline over time. InMalaysia, following an initial tightening, authori-ties in 1998 relaxed disclosure requirements, allow-ing disclosure to be less frequent (semi-annual);however, the regulatory reporting was not

changed.60 New bank laws (under preparation inThailand) or regulations (in Korea and Indonesia)also require banks to report to the public their fi-nancial statements more frequently (mostly quar-terly) or make their requirements more explicit.Consolidated financial reporting and disclosure bygroups will also enhance transparency; in Koreathis is promoted by the consolidation and concen-tration of supervisory functions. In the Philippines,banks listed on the stock exchange are now re-quired to disclose to the public on a quarterly basiskey indicators on their soundness.

The countries with blanket guarantees intend toremove them as soon as feasible.All countries in-

44

Table 13. Time Period for Overdue Criteria for Interest Suspension and Loan Classification

Period Overdue forCountry Interest Suspension Substandard Doubtful Loss

IndonesiaOld1 1–12 months 1–12 months2 21 months3

New 3 months 3 months 6 months 9 months

KoreaExisting Immediately when past due Normally not classified until 3 months

past due unless declared bankruptProposed No changes currently proposed 3 months 3–12 months 12 months

MalaysiaOld 6 months 6 months 9 months 12 monthsNew4 6 months 3 months 6 months 9 months

Philippines5 3 months 3 months (unsecured) . . .6 6 months (unsecured)

ThailandOld 6 months7 6 months (unsecured) Over 6 months Over 6 months

(unsecured)12 months (secured) Over 12 months Over 12 months

(secured)New 3 months 3–6 months 6–12 months Over 12 months

Source: IMF; national authorities.1Varies by type of credit and installment period.2Credit exceeds overdue criteria for substandard but is considered collectible and the value of collateral is not less than 75 percent or credit cannot

be collected, but value of collateral not less than 100 percent.3Refers to 21 months after a credit has been classified as doubtful and there is no repayment.4Effective March 1999.5New rules issued October 1997, which tightened overdue criteria for classifying loans depending on number and amount of arrearages, refer only to

installment loans.6A loan previously classified as substandard in the last examination is reclassified as doubtful if principal has not been reduced by atleast 20 percent

during the preceeding 12 months.7Effective January 1998 irrespective of collateral; previous limit (since July 1995) was 12 months for secured loan.

60Banking institutions were given an option of reporting non-performing loans using either the standard of three months or sixmonths past due. Of the 78 financial institutions, 21 (representing46 percent of the system’s loans) have retained the three-monthnonperforming loan classification criteria.

Regulations Governing Creditors

tend to move from a blanket guarantee to a limitedcoverage of all depositors.61 There are major dan-gers, such as deposit runs, in removing the guaranteebefore the financial system is sound, however. For alimited scheme to work, a number of preconditionshave to be met, including the banking system’s re-turn to solvency and profitability, the adoption ofmechanisms to deal with the exit of individualbanks, and the restoration of public confidence. Thepublic should be provided ample notice of the re-moval of the guarantee and detailed information onthe limited guarantee scheme that replaces it. The

lifting of the blanket guarantee, when it occurs,should be a nonevent.

Regulations Governing Creditors

Efforts are being undertaken to address weak-nesses in the operation and supervision of interna-tional lenders, as they have been identified duringthe Asian crisis. Initiatives include work on a newinternational financial architecture, the FinancialStability Forum, enhanced financial sector surveil-lance by international financial institutions, andwork by the Basel Committee on Banking Supervi-sion, whose Supervisory Lessons to be Drawn fromthe Asian Crisisproposes a review of various ele-ments of the existing capital rules and of the BaselCommittee’s guidance on country risk. (See BaselCommittee on Banking Supervision, 1999.)

45

Table 14. Loan Provisioning Requirements: Comparative Information(In percent)

Country Unclassified Standard Special Mentioned Substandard Doubtful Loss

IndonesiaOld1 .5 n.a. 102 502 1003

New 1 5 15 50 100

KoreaExisting .5 23 20 754 1004

ProposedNo changes currently proposed

MalaysiaOld5 0 0 206 506 1006

New5 0 0 206 506 1006

PhilippinesOld 0 0 07 50 100New8 0 5 259 50 1009

ThailandOld 0 0 1510,11 10010 10010

New12 1 2 2010 5010 10010

Source: IMF.1Based on uncollateralized portion.2Effective at the end of 1996 for Substandard and 1993 for Doubtful and Loss.3Classified as precautionary loans.4That portion of a loan classified doubtful or loss that is fully secured will normally be classified substandard to the extent of the market value of the

collateral.5Effective 1998 general provision increased from 1 percent to 1.5 percent of total outstanding loans (including interest), net of interest in suspect and

specific provisions.6Provision computed against uncollateralized portion.7For collateralized; 25 for uncollateralized.8Effective October 1997 a general provision of 2 percent on gross loan portfolio to be phased in through October 1999.adopted.9For both collateralized and uncollateralized.10Provision computed against uncollateralized portion.11Since June 1997.12Stricter criteria for secured loans.

61Korea’s limited deposit insurance scheme, which existed be-fore the crisis, is scheduled to replace the current blanket guaran-tee as of the end of the year 2000. Thailand did not have a depositinsurance scheme and is now working on the design of such ascheme to be introduced by special legislation.

Financial sector reforms have been at the core ofthe IMF-supported programs for the Asian crisis

countries. The focus was on the financial sectorbecause (1) distress and weaknesses in the sectorwere widely perceived as a major cause of the crisis;(2) reestablishing banking system soundness wascrucial to restoring macroeconomic stability;(3) restoring confidence in banks was essential forthe return of funding into the financial system; and(4) the crisis generated demands for the authoritiesto address the causes of the crisis by carrying outmajor reforms, thus providing impetus for imple-menting reform that in some cases had been plannedfor years. In the Philippines, the structural compo-nent in the program was smaller, because majorshortcomings of the financial system had been ad-dressed earlier. Malaysia’s strategy, although differ-ent in some respects from the IMF-supported pro-grams in the crisis countries, also had a major focuson financial sector reforms.

While the magnitude of the financial sector criseswas larger than anything experienced before, IMFstaff was able to draw on its experience in recentyears both from actual crisis involvement and analyt-ical work.62 Although many of the events in the un-folding crisis were familiar, differences in key ele-ments added new dimensions to the crisis requiringparticular care in choosing the restructuring ap-proach. Such elements were the structure of the bank-ing and corporate sectors, business practices, limita-tions in the legal systems, and the authorities’strongpreferences for certain institutional arrangements.The magnitude of the crisis and the speed at which itdeveloped required immediate responses. Thus, thedevelopment of strategies and policies involved ex-tensive discussions among the authorities and IMFstaff, and new approaches and procedures needed tobe developed, taking into account the characteristicsand constraints of each country at that particularpoint in time, while maintaining a certain amount ofuniformity of policies among the countries.

The reform strategy for each country was incorpo-rated in letters of intent and memoranda of economicand financial policies, both of which spell out the de-tails of the IMF-supported programs. These strate-gies included measures to ensure the exit of nonvi-able institutions, the strengthening of those thatremained in operation, the restructuring of the cor-porate sector, and the adoption of institutional re-forms to help prevent future crises. Box 13 summa-rizes the measures included in letters of intent andmemoranda of economic and financial policies ofthe countries that received IMF financial support.From the onset, these documents included both im-mediate actions and broad outlines of the medium-term strategy for restructuring and its sequencing.Given that the precise nature and timing of future ac-tions could not be determined at the outset, the pro-grams left sufficient flexibility in implementation.Subsequently, they were refined in the course of re-views as new information became available and asrequired by the evolving circumstances in eachcountry. The market sensitivity of certain actions re-quired in the initial stages of the program ruled outthe inclusion of such actions in the letters of intent.

Several issues arose in adapting complexmedium-term bank restructuring strategies to theformat and conditionality of IMF-supported pro-grams. IMF conditionality has usually been quantita-tive and strictly time-bound, involving actions underthe control of the authorities. In the case of bank re-structuring, however, conditionality has to be setcautiously, as the process involves steps that are sel-dom amenable to measurement, often take longerthan planned, are not directly under the control ofthe authorities, require legal steps to be adhered to,and involve negotiations between different parties inthe public and private sector. Moreover, since re-structuring actions have a significant impact on pri-vate property and wealth, they must be undertaken ina manner consistent with each countries’legal andjudicial framework.

The timing and pace of reforms requires a deli-cate balance between short-term IMF conditionalityand the medium-term nature of financial sector re-structuring. To maintain momentum and credibility,

VII Financial Sector Reforms in IMF-Supported Programs

46

62See, for example, Sundararajan and Baliño (1991), Lindgren,Garcia, and Saal (1996), Alexander and others (1997), Enoch andGreen (1997), and Folkerts-Landau and Lindgren (1998).

Financial Sector Reforms in IMF-Supported Programs

reforms must proceed rapidly, but the complexityof the process and country-specific constraints haveto be taken into account. For instance, rapid recapi-talization of financial institutions was a desirablegoal in principle, but in reality, too rapid a pacewould have meant that necessary concomitant mea-sures, such as operational restructuring and banks’search for private capital, would not have been fea-sible and that banks would have had to cut creditseven more drastically, thus aggravating the crises.Thus, the recapitalization was phased and strictlymonitored under the IMF-supported programs.Similarly, in Korea, policies to address the exces-sive maturity mismatches between foreign ex-change assets and liabilities had to be phased in,taking into account the difficulty in convertingshort-term to longer-term foreign financing at thetime. To introduce meaningful benchmarks for

measures like privatization, which involves com-plex negotiations with private parties, is even morechallenging.

Detailed information on the financial condition ofindividual banks was necessary in order to assess thesituation of the system, design a restructuring strat-egy, and monitor compliance with the program.Generally, it is impossible to deal with a systemicbanking crisis efficiently without access to bank-by-bank data. IMF staff had access to detailed supervi-sory data for individual banks in the three crisiscountries. Such access was necessary to evaluate thequality of the restructuring policies and to monitortheir implementation. In Malaysia and the Philip-pines, access to such data was more restricted. Han-dling of bank information had to be done accordingto agreed-upon procedures and in accordance withbank secrecy laws in each country.

47

Financial sector measures included in IMF-supported programs in Indonesia, Korea, Thailand,and the Philippines included:

Measures to stabilize the system

• Provide liquidity support at penal rates and subject toconditionality (Indonesia, Korea, and Thailand).

• Introduce a blanket guarantee (Indonesia andThailand).

• Cap deposit rates to reduce the ability of weak banksto capture deposits and further weaken the system(Indonesia and Thailand).

• Identify and close fundamentally unsound financialinstitutions. These included commercial banks (In-donesia), commercial and merchant banks (Korea),and finance companies (Thailand).

• Require owners of closed institutions to lose theirstakes in these institutions (Indonesia, Korea, andThailand).

• Share losses of closed finance companies with credi-tors; restructure some depositor claims to longer ma-turities (Thailand).

Measures to restructure the financial sector

• Establish a restructuring agency (Indonesian BankRestructuring Agency, IBRA, in Indonesia).

• Complete diagnostic reviews of financial institutions(Indonesia and Korea).

• Tighten loan classification and loan-loss provision-ing rules (Indonesia, Korea, Thailand, and thePhilippines).

• Allow for full tax deductibility on income for loan-loss provisioning (Indonesia, Korea, and Thailand).

• Establish a transparent timetable for banks to meetcapital adequacy requirements (Indonesia and Korea)or provisioning requirements (Thailand).

• Intervene in insolvent banks (all countries).

• Agree on memoranda of understanding between un-dercapitalized banks and the authorities to specify atimetable for raising capital to meet capital adequacyrequirements and attain performance benchmarks(Indonesia, Korea, the Philippines, and Thailand).

• Issue guidelines on the modalities for the use of pub-lic funds to recapitalize banks (Thailand) and to pur-chase nonperforming loans from private institutions(Indonesia and Korea).

• Issue guidelines for stricter bank licensing(Thailand).

• Take steps to privatize nationalized banks (Indonesia,Korea, and Thailand).

Measures to reform the institutional framework

• Enact legislation to enhance the operational indepen-dence of the supervisory authority (Korea) and cen-tral bank (Indonesia).

• Take steps to strengthen bank supervision (Indonesiaand the Philippines).

• Improve accounting, disclosure, and auditing stan-dards (Korea, the Philippines, and Thailand).

• Issue strengthened regulations regarding connectedlending, liquidity management, foreign currency ex-posure, and large exposures (Indonesia and Korea);cross guarantees were also to be eliminated for thetop 30 chaebols in Korea.

• Introduce a new bankruptcy law (Indonesia andThailand).

Box 13. Financial Sector Restructuring Measures in IMF-Supported Programs

While the Asian crisis showed that the IMF wouldneed to play a central role in assisting the authoritiesin the management of the initial crisis and in the de-sign of the overall restructuring strategy, it alsodemonstrated the need for close cooperation withother multilateral agencies, particularly the WorldBank. Cooperation with the World Bank was closefrom the beginning in all countries, although the di-vision of labor differed. In particular, the urgent na-ture of many of the tasks and the heavy demand onIMF staff resources required very close collabora-tion between the staffs of both institutions and flexi-bility in their approach to the division of labor.

The IMF assumed the lead role in the three crisiscountries and the World Bank made important con-tributions in specific aspects of program formulationand implementation. The IMF relied on its capacityto develop programs quickly and develop the link-ages between macroeconomic stability and financialsector soundness. The World Bank provided exper-tise, and financing, to assist the authorities in pro-gram implementation and institution building, in-creasing its role in the crisis countries over time. Inall crisis countries, the staffs of the IMF and WorldBank have cooperated closely from the early stages,taking into account each other’s views in the pro-gram discussions with the authorities—which oftenincluded staff of both institutions attending meetingswith the authorities on financial restructuring issues.Also, World Bank staff took the lead in the area ofcorporate restructuring and nonbank financial insti-tutions. Work on strengthening legal and regulatoryframeworks has been done jointly.

An important source of difficulties was related tothe design and implementation of the immediate stepsrequired to stabilize and restructure the financial sys-tem. In all three crisis countries, letters of intent werenegotiated quickly and incorporated immediate stepsto stabilize the financial system. These measures werenegotiated without full World Bank involvement, al-

though they included components for which theWorld Bank was expected to take the lead in imple-mentation. This led to coordination problems initially,further complicated by the large number of depart-ments involved in both institutions. In response, theWorld Bank created its Special Financial OperationsUnit to bolster its capacity to quickly field staff thatcan participate in the initial rounds of program nego-tiations in crisis countries. As a result of their efforts,early problems in coordination have been solved.

In addition to financial support, the IMF had to as-sign a large number of staff to the five countries dis-cussed in this paper. This was particularly the casefor the three crisis countries. During fiscal years1998 and 1999, staff from the IMF’s Monetary andExchange Affairs (MAE) Department and headquar-ters-based consultants spent some 10 staff years inthe field on missions to Indonesia, Korea, and Thai-land. These visits were often made in parallel with oras part of Use of Fund Resources and Article IV mis-sions. They were also done in conjunction withWorld Bank missions. Moreover, the IMF placedresident banking supervisors in Indonesia and Thai-land and organized several expert visits to Korea.Participation of financial sector experts in missionsto Malaysia and the Philippines has been less intensebut has also required substantial IMF staff involve-ment, more recently as part of the IMF’s Asia andPacific Department and World Bank missions. In ad-dition to mission-related work in Washington, MAEstaff and Washington-based consultants have spent aconsiderable part of their time at IMF headquartersfollowing events in all the countries, providing fol-low-up comments and assistance in implementation,and coordinating with other institutions. This is, ofcourse, in addition to IMF staff resources that otherIMF departments (notably the Asia and Pacific De-partment, or APD) had to commit to financial sectormatters in the context of the IMF’s day-to-day activ-ities and surveillance work.

48

VII FINANCIAL SECTOR REFORMS IN IMF-SUPPORTED PROGRAMS

The Asian experience with bank restructuringhas already produced valuable lessons, many

of which will evolve as the process continues. Inparticular, many “conclusions” at this time are actu-ally interim assessments, as it will take severalyears before the restructuring of the financial andcorporate sectors will be completed and the fulleconomic implications of different measures be-come apparent.

More than any other recent financial crisis, theone in Asia has highlighted the linkages between fi-nancial sector soundness and macroeconomic stabil-ity. Highly leveraged corporate sectors with largeamounts of unhedged foreign currency debt, much ofwhich is short-term, and domestic bank borrowing,fed by large capital inflows during years of excep-tional economic growth and exchange rate stability,created major vulnerabilities. The crisis highlightsthe danger that formally or informally pegged ex-change rates may lull investors into ignoring cur-rency risks (see Figure 8). Following the shocks tomarket expectations caused by exchange rate deval-uations and widespread doubts about private sectorsolvency, the size and speed of the impact on the fi-nancial systems was unprecedented. Foreign bankscut their credit, asset prices collapsed, leading tomajor wealth losses, and real demand contractedsharply. The close integration of the Asianeconomies in world financial markets helped tospread the crisis to other countries in the region andto the rest of the world. Close attention should bepaid to prudential supervision of foreign currencyexposures and risks, especially when exchange rateregimes are inflexible.

Why did the financial sector crises take everyoneby surprise? There were clear signals of overheating,such as prolonged rapid credit expansions, assetprice inflation, and overcapacity in key sectors, al-though the underlying deterioration in banks’loanvalues and capital adequacy ratios were not yet re-flected in their balance sheets. Meanwhile, thebuildup of corporate indebtedness, unsustainablebanking practices (such as directed, connected, andinsider lending) and weaknesses in prudential regu-lation and supervision were known to policymakers

and market participants. The fact that all these fac-tors were largely overlooked by most private sectorand published official analyses—both at home andabroad—was probably related to the long-runningsuccess of those economies. In any event, greaterdisclosure of macroprudential and microinstitutionalindicators and greater transparency of monetary andfinancial policies could have strengthened marketdiscipline and policy effectiveness and could havehelped to expose some vulnerabilities much earlier.

Compared with the three crisis countries,Malaysia and the Philippines had been pursuingpolicies before the crisis that clearly lessened thedamage. Both countries, which had undertaken bankrestructuring and structural reforms in the 1980s,avoided a full-blown crisis. Malaysia’s traditionalpolicies of limiting short-term foreign borrowing,encouraging foreign direct investment inflows, andrelying on equity capital prevented the corporatesector from building up the large unhedged foreignexchange exposures and very high debt equity ratiosthat were so damaging in the crisis countries. Thesepolicies notwithstanding, Malaysia also faced sub-stantial financial sector distress. In general, the crisishighlights the benefits of having developed money,bond, and equity markets. Developed capital mar-kets would reduce corporate leverage and improvecorporate governance; the reduced reliance on bankfinancing would make the system more resilient toshocks. The development of bond markets—espe-cially for government bonds—would also facilitatefinancial sector restructuring.

The first priorities in the crisis countries were tostop excessive central bank credit expansion to in-solvent institutions, stabilize the financial system,and prevent capital flight. To achieve this and to pre-vent bank runs, the governments needed to offerblanket guarantees for depositors and most creditors,close the worst institutions, and introduce crediblemacroeconomic stabilization and bank restructuringplans. These measures were successful in stoppingthe domestic deposit withdrawals, particularly afterthe credibility of the guarantees had been tested, butwere less effective in securing rollover of foreign li-abilities. To achieve the rollover, other measures,

VIII Conclusions and Lessons

49

such as the debt renegotiation in Korea, wereneeded. The experience of the crisis countries sug-gests that in a systemic crisis, a blanket guarantee,rather than a limited deposit guarantee, is needed torestore confidence in the financial system.

While closing the most insolvent institutions wasconsidered necessary, it also became clear that anyclosing of financial institutions and sharing of losseswith private sector creditors is an extremely difficulttask to manage. Closing deeply insolvent institutionsprovides a way to cease central bank support, allowsloss sharing with creditors, removes excess interme-diation capacity, and frees resources to deal with re-maining institutions. Decisions to close may have torely on liquidity “triggers” until banks’loan-lossesare recognized in their accounts. Any closing of fi-nancial institutions must be accompanied by a well-managed information campaign to support the pol-icy, explain the reasons for bank closings, andreassure the public.

In all countries, except initially in Indonesia, thecentral banks were able to sterilize their liquiditysupport to individual banks, since deposits with-drawn from weak banks were largely deposited inother domestic banks perceived as being safer. Toabsorb excess liquidity of the latter and stem capitaloutflows, interest rate increases for short-term cen-tral bank instruments were necessary. This recyclingwas successful in all countries, except in Indonesia,where economic and political turmoil made the situ-ation unmanageable during the first six months.Nonetheless, in all countries, exchange rates initiallydepreciated sharply due to concerns about credit-

worthiness and solvency of domestic counterpartiesand to uncertainties about policy implementation.

A slowdown in real credit growth to the privatesector should not have come as a surprise. In addi-tion to a sharp fall in credit demand, owing to over-capacity in the real sector and bleak short-term eco-nomic prospects, bankers became more cautious, asborrowers’creditworthiness deteriorated, reflectingin particular the large foreign exchange componentof their debt. This slowdown seems to be more ex-plained by structural factors than by an excessivelytight monetary policy. Any policies to stimulate banklending should be used cautiously so as not toworsen banks’conditions; for example, excessivelylow interest rate margins will be deterrents to newlending and further undermine banks’solvency andrecapitalization. Lending cannot be expected to nor-malize until the corporate sector has regained itsprofitability and repayment capacity and banks’cap-ital bases have been restored.

In broad terms, the strategies for systemic restruc-turing have sought to restore the financial systems tosoundness as soon as possible. The process involvesthe introduction of the necessary legal, institutionaland policy frameworks for dealing with nonviablefinancial institutions, strengthening viable ones, andresolving value impaired assets in the system. Sys-temic bank restructuring is a complex multiyear mi-croeconomic process, which in the aggregate hasmajor macroeconomic effects. No two situations arethe same, and specific measures, pace, and effective-ness of implementation have varied between the dif-ferent countries depending on the specifics of the

50

VIII CONCLUSIONS AND LESSONS

Indonesia (right scale)Thailand (left scale)

Malaysia (left scale)

Philippines (left scale)

Korea (right scale)

0

10

20

30

40

50

60

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

1993 1994 1995 1996 1997 1998


Figure 8. Nominal Exchange Rate(National currency per U.S. dollar, end of period)

Conclusions and Lessons

problem, laws and institutions in place at the outset,and the sometimes strong preferences of the authori-ties for certain solutions. For example, there havebeen major differences in the ability of each govern-ment to endow official agencies involved in the re-structuring with sufficient operational authority, fi-nancial independence, protection from lawsuits, andmeans to attract necessary expertise.

Key principles for the restructuring process in thethree crisis countries have included strict criteria toidentify viable and nonviable institutions and to re-move existing owners from insolvent institutions.Private capital injections have been encouraged orrequired under binding memoranda of understandingschemes that also foster operational restructuring. Tocomplement domestic equity investment in banksand to bring in international expertise, the three cri-sis countries liberalized their foreign ownershiprules for the financial sector. To encourage new pri-vate equity investment, public solvency support alsohas been offered under strict conditions.

Successful implementation of systemic bank re-structuring demands that national authorities takefull responsibility of all the aspects of the program.Given the microeconomic nature of financial re-structuring and the need for well-coordinated inter-play between so many different government institu-tions, the process can be successful only if theauthorities themselves take full control of the imple-mentation process. However, because systemicbanking crises occur very infrequently and requireskills not readily available in most countries, accessto international expertise can be particularly helpful.Thus, the IMF, the World Bank, and other interna-tional and bilateral agencies have provided adviceand shared experiences from other countries thathave dealt with similar problems. The institutionshave also helped countries obtain the necessary ex-pertise from the public and private sectors to assistin the restructuring.

A realistic valuation of financial institutions’as-sets is essential to measure net worth, yet extremelydifficult in cases of severe corporate and financialdistress. The valuation of nonperforming loans isparticularly hampered by the lack of clear marketvalues and continuously changing economic condi-tions. To better support the valuation process, allcountries tightened their rules for loan classification,loss provisioning, income recognition, and collateralvaluation, and have substantially strengthened theirsupervisory scrutiny of compliance with such rulesby bankers and auditors. Asset valuation is a key ele-ment in computing banks’capital adequacy require-ments, which have been the basis for their recapital-ization plans. In this connection, all the countriesseek to bring their capital adequacy requirementsinto full compliance with international standards by

the end of the restructuring process. In doing so,banks have been given time, according to transpar-ent rules, to gradually restore their severely erodedcapital bases. Recent developments suggest thattransparency in loss recognition practices has helpedin the issuance of new private bank equity.

Management of nonperforming loans and othervalue-impaired bank assets is one of the most criticaland complex aspects of bank restructuring. There isno single optimal strategy for all circumstances butrather a combination of options that may vary overtime and for each bank, depending on factors such asthe nature of the problem assets, their overall sizeand distribution, and legal and governance con-straints. The strategy will need to consider the speedof disposition of the assets, whether or not to use acentralized process, which also involves ownershipchoices.

Countries have taken different views on the roleof asset management companies. Indonesia, Korea,and Malaysia have opted for centralized public assetmanagement companies that buy assets from privatebanks to help banks clear their balance sheets. Thai-land has aggressively liquidated the impaired assetsof closed institutions through a central agency butdoes not permit public sector purchases of impairedassets from private banks; the Thai authorities areinstead encouraging each bank to establish its ownasset management company. The key issue in assetpurchases by an asset management company is real-istic valuation, so as to ensure that it does not be-come a tool for the indirect bailout of existing share-holders, thus undermining the incentives for privatesector recapitalization and proper governance ofasset management companies and banks. The sale ofbanks’ assets to an asset management companyforces immediate recognition of the value of theloan. This may deter such sales in cases where bankshave been carrying these loans at inflated values.

The use of vast amounts of public resources hasbeen necessary to help restructure the financial sys-tem. This affects medium-term fiscal sustainabilityin the three crisis countries. An estimate of the exactcost of the crisis is not possible, as the costs are stillemerging and the corporate debt restructuringprocess is far from completed. The outlays initiallytook the form of central bank liquidity support, buthave been gradually identified as solvency supportand been refinanced by public or publicly guaran-teed debt. Only the carrying costs of the latter havebeen brought into the government budgets so far.The ultimate public sector costs will be reduced byproceeds from asset sales and bank privatizations.Moreover, more rapid and sustained economicgrowth as a result of a restructured, more efficient,and profitable financial system will yield additionaltax revenue.

51

The problems in the financial sector also reflectedprofound problems in the corporate sectors. Solvingthe banking and the corporate sector crises thereforemust go hand in hand. Resolving banking problemsrequires not only bank recapitalization, but alsobroad-based corporate debt restructuring. At thesame time, resolving corporate sector turmoil, in turnrequires properly functioning banks as a counterpart.Widespread corporate weaknesses and insolvenciesare much more difficult and time-consuming to dealwith than bank restructuring. Bankruptcy and otherlaws, which were either not in place at the onset ofthe crisis or inadequate, as well as judicial systems,are now being reformed to speed up corporate debtrestructuring. Nevertheless, lags in corporate reformare slowing progress in bank restructuring.

Bank restructuring provides a key lever for corpo-rate restructuring. Tighter and better enforced pru-dential regulations can induce the financial restruc-turing of corporations, especially those that arehighly leveraged. Moreover, banks can play a leadrole in inducing corporate restructuring, particularlyif they have strong government support to do so, asis the case in Korea.

The IMF programs designed to deal with the prob-lems of crisis countries in Asia have centered onstructural reforms of the financial sectors since it wasclear that macroeconomic measures alone would beinsufficient. This represented a major departure froma traditional IMF-supported adjustment program.Program design and implementation allowed IMFstaff to share with the authorities broad experiencesfrom systemic crisis management in other countries.At the same time, the magnitude, depth, and localcircumstances of each country required quick and in-novative responses and a major deployment of IMFstaff and experts in close cooperation with the WorldBank. The structural components and conditionalityof IMF- and World Bank-supported programs wereclearly instrumental in implementing the authorities’restructuring programs.

The different country experiences have providedIMF staff with important lessons in actual crisismanagement. The crisis experiences have confirmedthat all major elements of the IMF’s Toward aFramework for Financial Stability, prepared in 1997and published in 1998, remain valid (see Folkerts-Landau and Lindgren, 1998). The crisis has againhighlighted the main vulnerabilities identified in thatpaper: problems of valuing bank loans (identified as“the Achilles’ heel” of effective corporate gover-nance, market discipline, and official oversight) andtherefore banks’net worth; design of lender-of-last-resort facilities; deposit guarantee schemes and exitprocedures; and importance of supervisory authorityand capacity. Areas in which that paper could be ex-panded would be more in-depth analysis of the

macroeconomic environment as a source of financialsector vulnerability and the relationship between thefinancial and real sectors.

Could the crisis have been prevented? In eachcountry, alternative courses of policy that wouldhave been preferable can be identified—but this is togive hindsight too much of a role. In the financialsector itself, more transparency of macro- and mi-croeconomic data and policies clearly would havehelped in bringing matters to a head earlier, whichcould have lessened the depth of the crisis. Policiesto foster corporate governance and lower corporateleverage would also have helped. Better regulatoryand supervisory frameworks would have helped, forexample, in detecting and correcting problems suchas excessive real estate lending and the proliferationof financial institutions. However, even if the BaselCore Principles (of April 1998) had been in place,bank supervisors probably would have been unableto take restrictive action, since the underlying prob-lems were masked by the economic boom. For simi-lar reasons, there was no way for fiscal policies toanticipate the size of the fiscal contingent liabilitiesbuilding up in the financial systems. In particular,foreseeing the sudden and massive erosion of loanand asset values that took place once market senti-ment changed and the exchange rates collapsedwould have been very difficult.

Comprehensive reforms of legal, institutional, andadministrative frameworks have been initiated withthe aim of introducing international standards andbest practices. Such reforms encompass moderniza-tion of financial sector laws and prudential regula-tions, including operational procedures for exit ofproblem banks. Reforms also include a strengthen-ing of supervisory powers, procedures, and capabili-ties with the aim of bringing about better risk man-agement in banks. Another key area of reform is newlaws and procedures for corporate bankruptcy andgovernance.

The above reforms are facilitated by the broad-based international efforts under way to implementnew standards, codes, core principles, and best prac-tices. International initiatives have been undertakento improve financial sector architecture, surveillanceover national and international financial markets,dissemination of data, macroeconomic vulnerabilityindicators, and prudential regulations and supervi-sion of international lenders. This last initiative, ledby the Basel Committee on Banking Supervision,aims at revising, among others, solvency require-ments and country risks.

National authorities may also draw other lessonsfrom the crisis in terms of countercyclical prudentialpolicies and more supportive macroeconomic poli-cies. Banks’ net worth, including capital and re-serves, should be built up in times of economic

52

VIII CONCLUSIONS AND LESSONS

Conclusions and Lessons

booms so that a cushion is available to deal with in-evitable downturns. Such prudential policies areoften referred to as countercyclical, and could in-clude a strengthening of requirements regarding liq-uid assets, collateral margins, capital adequacy ratios,and general loan-loss provisioning in good times—when the future bad loans are extended. It shouldalso be stressed that prudential policies are not sub-stitutes for a proper mix of macroeconomic policies.

The experience of the crisis countries illustratesonce more the importance of prompt and decisiveaction to deal with banking problems, including pre-emptive bank restructuring actions, and the dangersof waiting for the situation to reverse itself.Preven-tive action, notably in prudential regulation and su-

pervision and in the form of more transparency,would have improved bank governance and marketdiscipline and helped prevent weaknesses frombuilding up. Early action also could have reducedthe magnitude of the portfolio losses and the needfor liquidity and solvency support. Conversely, oncecredible programs were put in place with broad do-mestic and international support, the crisis becamemanageable and provided a foundation for economicrecovery: the scale and complexity of the problemswould have made it difficult to address them withoutsuch support. This also suggests the importance ofcountries obtaining international assistance early inthe process, in which the IMF and the World Bankplay key roles.

53

In the decade prior to the emergence of the Asiancrisis, Indonesia had undergone rapid develop-

ment, with per capita income doubling from 1990 to1997.63 This resulted from a significant diversifica-tion of the economy away from a dependence on oiland gas to a wider export basket consisting of pri-mary products, textiles, and light manufactures. Eco-nomic growth was supported by a stable macroeco-nomic environment, characterized by balancedbudgets and low inflation policies geared to main-taining a stable real exchange rate.

Indonesia has a tradition of a liberal capitalregime. Restrictions on both inflows and outflowswere largely relaxed in the mid-1970s. Nevertheless,until the early 1990s capital inflows played only aminor role, given a relatively undeveloped financialsystem, while the stable macroeconomic environ-ment avoided any large capital outflows.

Background

In the 1990s, Indonesia significantly liberalized itsdomestic economy but direct and indirect state influ-ence still remained very important. In the financialarea, less government control led to a significant in-crease in the number of banks, often associated withspecific industrial groups. A stock exchange andother financial institutions complemented the sys-tem. These developments notwithstanding, statebanks continued to play a major role in the system.

Macroeconomic Setting

Prior to the crisis, Indonesia had been strugglingwith significant capital inflows. Bank Indonesia had,on several occasions, widened the exchange rateband, each time leading to a maximum appreciationof the rupiah within the band. In response to the realappreciation of the currency, exports in late 1996and early 1997 showed rapid signs of decline.

The crisis in Indonesia was essentially triggeredby contagion from Thailand, especially after Thai-land floated its currency, the baht, in July 1997. Con-tagion occurred despite the fact that Indonesia hadstronger macroeconomic fundamentals than Thai-land, particularly as they pertained to exports and thefiscal balance. Following a widening of the rupiah’sband in July, the currency was floated in August1997. After a temporary reprieve, exchange marketpressures heightened in September and October, bywhich time the rupiah had fallen by more than 30percent since July, the fastest depreciation amongthe crisis countries. On November 5, 1997, the IMFapproved a three-year Stand-By Arrangement withIndonesia, equivalent to $10 billion; additional fi-nancing was committed by other international finan-cial institutions and contingent credit lines fromother countries. The initial response to the programwas positive, but it proved short-lived, and the ex-change rate fell again precipitously in December1997. A strengthened program was announced inJanuary 1998 but political difficulties in the run-upto, and after, the presidential election led to severeeconomic disruption in the economy. In June, the ru-piah hit an end-of-the-day low of 16,650 rupiahagainst the U.S. dollar.

In the aftermath of the May 1998 riots and the re-placement of the President, the macroeconomic pro-gram was modified and a limited degree of macro-economic stability was restored. The exchange raterecovered to 11,075 rupiah against the U.S. dollar bythe end of August 1998, and has stabilized at about8,000 rupiah per U.S. dollar since late 1998.

Characteristics of the Financial Sector

Rapid economic growth, a liberal capital accountregime, and regulatory changes have all contributedto substantial development in Indonesia’s bankingsector. The main characteristics of the system at theoutset of the crisis can be summarized as follows:

• Size and concentration.After relaxing bankentry, the number of banks increased from 111in 1988 to 240 in 1994. (The financial system inIndonesia remains dominated by commercial

Appendix I Indonesia

54

63This appendix draws on a draft by Charles Enoch andOlivier Frécaut.

Appendix I

banks; see Box 14.)64 A 1994 increase of mini-mum capital requirements from 10 billion ru-piah to 50 billion rupiah reduced the number ofpotential entrants, leaving the number of institu-tions almost constant until the crisis broke. Theseven state banks had combined assets account-ing for about 40 percent of the entire system, al-though their share of total bank lending declinedmarkedly after liberalization.65

• Ownership and entry. Liberalization increasedthe attraction of the financial sector to commer-cial and industrial concerns, and many of In-donesia’s large business conglomerates foundedone or more banks. The 10 foreign banks thatoperated in Indonesia obtained licenses in thelate 1960s. Since then and until 1999, the entryof foreign banks was limited through the re-quirement to form either joint ventures (with amaximum of 85 percent foreign ownership) orbuy shares of domestic banks on the Stock Ex-change where the maximum foreign holdingwas set at 49 percent.

• Areas of business activities. Domestic bankswere required to direct 20 percent of credit tosmall-scale business projects, and foreign bankswere required to lend 50 percent to export-oriented businesses, although these require-ments were often not met.

Weaknesses in the Financial Sector

Several issues related to ownership and businesspractice made the Indonesian financial sector vul-nerable to the shocks that were experienced in1997.


• Nontransparent ownership and portfolio prob-lems. Quantification of the extent of problemloans in Indonesia was difficult, given the largenumber of banks and the complex pattern ofcross holdings of equity and loans, which im-paired the transparency of reports.

• Loan concentration in the real estate sector.While difficult to detect in the data collected byBank Indonesia before the crisis, there had been

a sharp increase in real estate and property-related lending, which increased to about 20percent of total outstanding loans in early 1997.In Indonesia, the dangers of loan concentrationwere heightened by difficulties in seizing andrealizing collateral.

• Exposure of banks to market risks. From 1988,Bank Indonesia regulated banks’activities inother financial areas and limited banks’directinvolvement in leasing, venture capital, securi-ties trading, and investment management.Banks were, however, permitted to pursue suchactivities through the formation of subsidiariesoperating as nonbank financial institutions, andthey made frequent use of this option.

Prudential Regulation and Supervision

In the late 1980s and the first half of the 1990s, In-donesia improved banking regulation and supervi-sion. A comprehensive set of well-drafted and up-to-date prudential regulations was prepared and issued.An advanced version of the U.S.-inspired capital,asset, management, equity, and liquidity (CAMEL)rating system was put in place. All these regulationswere frequently updated and improved. Yet short-comings in the legal and regulatory framework re-mained, particularly in the areas of loan classifica-tion and establishment of an effective exitmechanism for failed banks. Even more important,enforcement of the regulations was a major problem,often owing to political interference.

Prudential regulation

• Loan classification and provisioning. Evenwhen problem loans were identified, loan clas-sification standards in Indonesia remained inad-equate because it was easy to restructure loansto reduce the size of reported portfolio prob-lems. Moreover, Indonesian standards allowed abank to reclassify loans back to performing sta-tus as soon as one payment was made, irrespec-tive of the anticipated future payment stream onthe loan. Banking supervisors, while recogniz-ing the drawbacks of these practices, had not fo-cused thoroughly on the extent of loan restruc-turing as an additional indicator of bankingsector soundness.

• Bank exit. No effective bank closure and exitregulation was in place. Instead, failed privatebanks were generally absorbed by Bank Indone-sia. In late 1996, a bankruptcy law for bankswas passed, but it was deficient because itgranted important rights to shareholders in theliquidation process and foresaw a liquidationprocess lasting several years.

55

64This discussion excludes 9,200 rural banks, which are limitedin the geographical area of operations, may only accept time andsaving deposits and extend loans, and hold 1–2 percent of totalbanking assets. Minimum capital for rural banks is 50 million ru-piah.

65Five state banks were included in a World Bank rehabilitationproject, which at the time imposed limits on their asset growth.

APPENDIX I

56

Box 14. Indonesia: Outline of Steps Toward Bank Resolution

The process of resolving the banking crisis can bebroken into nine steps.

Step 1.As of the end of June 1997, before the crisis,the Indonesian banking system consisted of the follow-ing categories:

Number Share of Assetsof (In percent,

Banks June 1997)

Private domestic banks 160 49.4With foreign exchange

(forex) license 80 43.9Without forex license 80 5.5

Publicly owned banks 34 42.3State-owned banks 7 39.8Regional development banks 27 2.5

Joint ventures and foreign banks 44 8.3

Joint ventures 34 4.7Branches of foreign banks 10 3.6

Banking system 238 100

Step 2.As part of the commitments included in theOctober 31, 1997, letter of intent, 50 banks identifiedas weak were subjected by Bank Indonesia to specificresolution measures, which included the closure of 16private domestic banks:


Banks June 1997)

Private domestic banks 41 24.1Closed (5 forex and

11 nonforex) 16 2.5Other resolution measures

(13 forex and 12 nonforex) 25 21.6Publicly-owned domestic banks 8 10

State-owned banks 2 9.6Regional development banks 6 0.4

Joint-ventures and foreign banks 1 0.2Joint ventures 1 0.2Branches of foreign banks 0 —

Total 50 34.3Banking system (before the

16 closures) 238 100

Step 3. On February 14, 1998, the Indonesian BankRestructuring Agency (IBRA) took over the surveil-lance of 54 banks, including 4 state banks subject to arestructuring plan under World Bank auspices, and50 private and regional development banks that hadborrowed from Bank Indonesia more than 200 percent

of their capital, and had a capital adequacy ratio below5 percent on adjusted end of December 1997 figures:


Banks June 1997)

Private domestic banks 37 11.2Forex 22 10.8Nonforex 15 0.4

Publicly-owned domestic banks 15 25.0State-owned banks 4 24.7Regional development banks 11 0.3

Joint ventures and foreign banks 2 0.5Joint ventures 2 0.5Branches of foreign banks 0 —

Total-BRA Banks 54 36.7Total-Non-IBRA Banks 168 63.3Banking system 222 100

Steps 4-Aand 4-B. On April 4, 1998, IBRAtook ac-tion against the 14 worst banks placed under its surveil-lance since February.

The seven largest borrowers from Bank Indonesiahad borrowed more than 2 trillion rupiah each and ac-counted together for over 75 percent of the total BankIndonesia liquidity support to the banking system.IBRA took control through suspending shareholders’rights and (apart from in the state bank) changing themanagement of the following banks. These wereknown as banks-taken-over (BTO).

Share of liabilities(In percent,

Name Category June 1997)

Bank Umum National (BUN) Forex 1.1Bank Dagang Nasional

Indonesia (BDNI) Forex 3.4Bank Modern Forex 0.3Bank Danamon Forex 2.2Bank Tiara Asia Forex 0.5PDFCI Joint Venture 0.4Bank Ekspor Impor

Indonesia (EXIM) State 7.7Total 15.6

Seven small banks in a particularly critical conditionhad borrowed from Bank Indonesia more than 500 per-cent of their equity and 75 percent of their assets. Thesebanks were frozen, equivalent to a closure, and theirdeposits transferred to designated state banks (BBObanks, or Bank Reku Operati, i.e., banks whose opera-tions are frozen):

Appendix I

57

Share of Liabilities(In percent,

Name Category June 1997)

Bank Surya Forex 0.15Bank Pelita Forex 0.11Bank Subentra Forex 0.08Bank Hokindo Nonforex 0.01Bank Istismarat Nonforex 0.02Deka Bank Nonforex 0.01Centris International Bank Nonforex 0.02Total 0.40

As a result, the total number of active banks in In-donesia was reduced from 222 to 215, of which 47were under IBRA’s, and 168 under Bank Indonesia’ssurveillance.

Step 5. On May 29, 1998, following relentless runsthat led Bank Indonesia to provide it with 32 trillion ru-piah of liquidity support, Bank Central Asia, the largestdomestic private bank (12.0 percent of the liabilities ofthe banking sector) was taken over by IBRA, the own-ers’ rights were suspended, and the management re-placed, and Bank Central Asia became the eighthBTObank.

The number of IBRAbanks went from 54 to 55 (40active, 8 BTO, and 7 frozen), but was shortly thereafterreduced to 53 when 2 nonforex IBRAbanks weremerged with another IBRAbank.

During the same period, two non-IBRAprivate do-mestic banks (one forex and one nonforex) weremerged with other non-IBRAbanks. At that stage, thedistribution of banks was therefore as follows:


Banks June 1997)

Private domestic banks 127 23.1Forex 68 21.7Nonforex 59 1.5

Publicly-owned domestic banks 41 67.5State-owned banks 7 44.6Regional development banks 27 1.9IBRA banks (BTO) 7 21.0

Joint ventures and foreign banks 43 9.4Joint ventures 33 4.7Branches of foreign banks 10 4.7

Total active banks 211 100

Step 6.On August 21, 1998, based on the results ofportfolio reviews, the authorities announced that threeof the banks taken over by IBRAon April 4 (Step 4-A),namely Bank Umum Nasional, Bank Dagang NasionalIndonesia, and Bank Modern would be frozen, and

their deposits transferred to designated state banks. Thenumber of active banks was to be reduced to 208.Danamon was to be recapitalized by the government toserve as a bridge bank, PDFCI and Bank Tiara Asiawere to be given a final opportunity for recapitalizationby their owners, and negotiations began with the own-ers of Bank Central Asia to reacquire their bank.

Step 7.On September 30, the authorities announcedthe intent to merge four banks (Bumi Daya, Bank Pem-bangunan Indonesia, or BAPINDO, BRI, and EXIM)into the newly formed Bank Mandiri. They also an-nounced a plan for the recapitalization of the statebanks.

Step 8. In the October 19, 1998 SupplementaryMemorandum of Economic and Financial Policy, theauthorities announced their intention to commence theliquidation of the 10 IBRAfrozen banks (Steps 4-B and6). The distribution of surveillance responsibilities forthe 208 active banks changed again, as only the BTOand frozen banks remained under IBRA’s responsibil-ity, while 11 other banks were returned to Bank Indone-sia for supervisory purposes (in advance of theprospective restructuring of the state banks).

Step 9. On March 13, 1999, the government an-nounced the results of the private recapitalization pro-gram; 74 banks, comprising 6 percent of total bankingsector assets, had over 4 percent capital adequacy re-quirement and were categorized as “A” banks. Ninebanks, comprising 12 percent of liabilities, had capitaladequacy requirements between 4 percent and –25percent, and were deemed eligible for recapitalization;seven banks, comprising 4 percent of liabilities, weretaken over by IBRA; and 38 banks, comprising 5 per-cent of liabilities, were closed, including 17 bankswith capital adequacy requirements worse than –25percent (the “C” category banks). After these mea-sures, the breakdown of the banking sector was asfollows:


Banks June 1997)

Private domestic banks 83 17.9Forex 30 4.8Nonforex 44 1.1Eligible for recapitalization 9 12.0

Publicly-owned domestic banks 45 71.8State-owned banks 7 50.3Regional development banks 27 2.1IBRA banks (BTO) 11 19.3

Joint ventures and foreign banks 40 10.3Joint ventures 30 5.1Branches of foreign banks 10 5.2

Total active banks 168 100

APPENDIX I

Prudential supervision

Despite the improvements in the supervisoryframework, serious implementation shortcomingsremained. These included:

• Substantial forbearance. Violations of pruden-tial rules were not appropriately sanctioned andnoncompliance was widespread.

• Ineffective on-site supervision. Onsite inspec-tions in Indonesia yielded limited additional in-sight into the actual number of problem loans,in contrast with the experience in other coun-tries, where these inspections usually found amuch higher number of nonperforming loansthan reported by banks.

• Insolvent banks remained in the system. Giventhe problems with bank closure, several knowninsolvent banks remained open. While the com-bined overall negative net worth of insolventbanks remained relatively small (about 0.5 per-cent of GDPin 1996), the situation createdmoral hazard problems.

Impact of the Crisis on BankPerformance and Initial Resolution Measures

The following provides a chronology of the devel-opments in the Indonesian banking crisis and sum-marizes the initial reactions.

From a Limited to a Systemic Banking Crisis:October–December 1997

When Indonesia requested IMF assistance in earlyOctober 1997, teams from the IMF, the World Bank,and the Asian Development Bank worked jointly toreview the condition of the banking sector to providesupport on financial sector issues. Based on datamade available by Bank Indonesia, the financialcondition of 92 of the 238 banks, representing 85percent of the assets of the banking system, wasevaluated. The depreciation of the rupiah, combinedwith a sharp shift in market sentiment, had alreadymade a serious impact on the banking sector. At thatstage, however, the banking sector did not show thecharacteristics of a systemic banking crisis despitedeposit withdrawals from some small banks. State-owned banks’weaknesses appeared manageable aspart of the fiscal adjustment, and most major privatebanks still posted comfortable cushions of positiveequity.

In the framework of the overall program, aimingat a swift macroeconomic recovery, the IMF and theIndonesian authorities agreed on a comprehensivebank resolution package consisting of:

• intensified supervision, including frequent anddetailed reviews, in addition to daily monitor-ing of key elements like liquidity and foreignexchange exposure for six of the country’slargest private banks (market share: 18.0 per-cent) in which some critical weaknesses hadbeen identified;

• rehabilitation plans, based on memoranda of un-derstanding or cease-and-desist orders for sevensmall private banks (market share: 0.7 percent)in which serious weaknesses, including under-capitalization, had also been identified;

• conservatorship for three small, severely under-capitalized private banks (market share: 0.1 per-cent), and for six insolvent regional develop-ment banks (market share: 0.4 percent), pendingthe results of discussions with the regional gov-ernment owners;

• transfer of the performing assets for two insol-vent state-owned banks (market share: 9.6 per-cent) to a third state-owned bank; merger of thetwo insolvent banks, and transformation of theresulting entity into an asset recovery agency;

• definition and implementation of rehabilitationplans for 10 insolvent private banks (marketshare: 3.0 percent) that had benefited from aBank Indonesia-sponsored and legally bindingrescue package prior to the crisis, acceleratingtheir return to solvency; and

• immediate closure of 16 small and deeply insol-vent private banks (market share: 2.5 percent),with protection limited to small depositors.

In total, the resolution package announced to thepublic included 50 banks, representing 34.3 percentof the banking system. This package included theclosure of the 16 small private banks. Since the re-maining 34 banks were not identified, however, thiscreated uncertainty among the public regarding thefate of all other banks. In the announcement, BankIndonesia indicated its readiness to provide protec-tion against runs through liquidity support for allbanks that remained open, while a comprehensiveaction plan to improve the institutional, legal, andregulatory framework was set in motion.

After a short-lived positive reaction to the recov-ery program, and an appreciation of the rupiah, theenvironment rapidly began to deteriorate again. Byearly December, a number of elements had com-bined to reduce public confidence in the bankingsector:

• the depreciation of the rupiah, high interestrates, and a slowdown in the economy took anincreasing toll on bank profitability andsoundness;

• rumors about the President’s health and his lastminute cancellation of a high-profile trip abroadcreated an atmosphere of political instability;

58

Appendix I

• the fast deterioration of the macroeconomic sit-uation in Korea added to the uncertainty; and

• the developments regarding Alfa Bank wereviewed as a sign that the authorities were notgenuinely determined to implement the programas agreed with the IMF.66

The deposit runs multiplied amid rumors that anew wave of bank closures was under preparation,and the segmentation of the interbank market inten-sified. By mid-December 1997, 154 banks repre-senting half of the total assets of the system hadfaced, to varying degrees, some erosion of their de-posit base. During December 1997, Bank Indone-sia’s liquidity support increased from 13 trillion ru-piah to 31 trillion rupiah, equivalent to 5 percent ofGDP. Insofar as the liquidity support, paid in rupiah,was needed by banks to meet reductions in dollar de-posits, in effect it served to fuel capital flight and,thus, the continuing depreciation of the exchangerate. In contrast to other crisis countries, efforts atsterilization were not successful, reflecting a loss ofmonetary control by Bank Indonesia.

Stabilizing the Banking Sector:January–February 1998

During January 1998, a sudden and deep deterio-ration in the economic environment took place, withthe rupiah heading into a free fall. The rate fell from4,600 rupiah per U.S. dollar at the end of December1997 to 15,000 rupiah per U.S. dollar in late January1998, with some trades even at 17,000 rupiah. TheIndonesian authorities signed a letter of intent withthe IMF on January 15, 1998, but continued evidentfailure to live up to commitments by the Indonesianauthorities led to further exchange rate depreciation,notwithstanding renewed discussions to bolster thecommitments made in the January 1998 letter of in-tent. When it became impossible for Indonesia topresent a credible recovery program to the IMF, thebanking sector problems turned into a full-fledgedsystemic crisis, with liquidity support from Bank In-donesia exceeding over 60 trillion rupiah (about 6percent of 1998 GDP), risk of hyperinflation, andcomplete financial sector “meltdown.”

On January 27, 1998, a new financial sector strat-egy was introduced, with the authorities’immediatepriority being to avoid a financial collapse and tostabilize the banking sector. The government an-nounced a three-point emergency plan. First, all de-

positors and creditors of all domestic banks were,henceforth, to be completely protected.67 Second,IBRA was established for a period of five years,under the auspices of the ministry of finance, to takeover and rehabilitate weak banks and manage thenonperforming assets of intervened banks. Third, aframework for handling corporate restructuring wasproposed. The impact of the announcement was im-mediate, with the exchange rate recovering to12,000 rupiah per U.S. dollar and appreciating fur-ther in subsequent days. Attempts were made toplace restrictions on banks’activities to mitigate themoral hazard effects of the blanket guarantee; for in-stance, deposit rates were capped at a specific pre-mium above the levels being offered by the best-runbanks.

In the following two months, efforts were made toreestablish monetary control by rationalizing BankIndonesia liquidity facilities and by developing ef-fective penalties to deter banks from seeking accessto these facilities. To better assert the authorities’control, surveillance over 54 banks (comprising 36.7percent of the banking sector; see Box 14, Step 3)that had borrowed heavily from Bank Indonesia wastransferred to IBRAon February 14. This includedfour state-owned banks (BAPINDO, Bank BumiDaya, BDNI, and Bank Exim), accounting for 24.7percent of the liabilities of the banking sector. IBRAexaminers were placed in those banks, and the bankswere required to sign memoranda of understandingsetting out the strengthened supervision under whichthey would operate.

While this initial operation was carried out rela-tively smoothly, its impact was much reduced by therefusal of the President to publicize it lest it trigger arenewal of runs. As a result, the initial workings ofIBRA were not apparent to the public, there wasconfusion as to the authorities’intentions, and themomentum generated by the January 27 announce-ments was largely lost. Even worse followed when,in late February, the President dismissed the head ofIBRA, at that time a highly respected finance min-istry official, and some of the staff seconded intoIBRA were returned to Bank Indonesia. The govern-ment itself drifted through the period up to the presi-dential election of 1998 and a prolonged debate en-sued over the possible introduction of a currencyboard arrangement. Liquidity support to the bankingsystem continued to increase, largely to meet contin-uing deposit withdrawals but also in the face ofwithdrawals of credit lines to domestic banks, as

59

66The President’s son, whose Bank Andromeda had beenclosed on November 1, was allowed to take over the small AlfaBank, which was immediately granted a foreign exchange licenseby Bank Indonesia, and to transfer into it most of his former ac-tivities, effectively reopening his former bank under a newname.

67The blanket guarantee was to last a minimum of two years,and at least six months’notice would be given before its termina-tion. Subsequently, the guarantee was retroactively applied to the16 closed banks and thereafter to the 9,200 rural banks.

APPENDIX I

well as emerging losses in derivatives businesses ina few banks, including one state bank with morethan 20 trillion rupiah (2 percent of GDP) of suchlosses. Altogether, by that time, liquidity support tobanks stood at more than 60 trillion rupiah (6 per-cent of GDP).

First Resolution Initiatives and New Shock:March–May 1998

The ensuing three months saw the authorities tak-ing their first initiatives to resolve ailing banks andthen facing a major new shock. With monetary con-ditions progressively being brought under control,the main focus turned to establishing the necessaryinfrastructure for handling the banking crisis: mak-ing IBRA operational, preparing the legal frame-work, obtaining better information on the financialcondition of the banks, and beginning to takeaction.

In early March 1998, Bank Indonesia announcedthe redesign of its liquidity support facilities. Untilthen, support had been provided through severalwindows—with the deterrent to usage, in theory,being highly punitive interest rates. Virtually nobanks, however, were actually paying interest,which was therefore routinely capitalized, causing arapid further expansion in outstanding liquidity sup-port. The new system involved a single liquidity fa-cility with interest rates, generally only a small mar-gin above market rates. The new focus was onnonmarket sanctions; any bank with borrowings out-standing for more than a week would have a specialBank Indonesia inspection, which would produce areport within a further week, increase restrictions onthe bank’s activities, and culminate in possible trans-fer to IBRA.

By late March 1998, the new IBRAmanagementteam was ready for more substantive interventioninto the most critical of IBRAbanks. On April 4, inits first major public action, IBRAtook over theseven banks (Box 14, Step 4, part A) that had eachborrowed more than 2 trillion rupiah (all but two ofthese had borrowed more than 5 trillion rupiah each,and two of them over 20 trillion rupiah); togetherthey accounted for about 72 percent of total Bank In-donesia liquidity support to the banking system. Thefocus at this stage was on liquidity, rather than sol-vency, to determine which banks should be inter-vened. This was partly because, in the absence of re-liable data on the solvency condition of the banks,liquidity criteria could serve as a proxy, and partlybecause of the urgent need to tackle the provision ofBank Indonesia liquidity support itself in order tostabilize monetary conditions.

For the six private banks among them, ownerswere suspended and managements removed; new

management was put in place through “twinning”arrangements with designated state banks.68 At thesame time, seven smaller banks (Box 14, Step 4, partB), comprising 0.4 percent of the banking system,which had each borrowed more than 500 percent oftheir capital, were closed; all deposits were trans-ferred over that weekend into a designated statebank, Bank Negara Indonesia. Great efforts weremade to ensure uniform application of objective cri-teria in the choice of both sets of banks, and therewas an intensive and professional public relationscampaign over the weekend to explain the moves tothe public. As a result, the moves were received fa-vorably in the markets; there were sporadic runs on afew of the acquired banks, but these diminished.These actions were a major step to demonstrate theauthorities’commitment toward bank resolution andto bring a revised IMF-supported program to theIMF’s Executive Board for consideration in lateApril.

The riots in mid-May of 1998 led to a depreciationof the recently stabilized rupiah and a further loss ofconfidence by both domestic and foreign investors.In the aftermath of the riots, there were massive runson Bank Central Asia, the largest private bank,which accounted for 12 percent of total banking sec-tor liabilities. Given the circumstances, support waseffected relatively smoothly. Bank Indonesia, in con-junction with two of the state banks, supplied over30 trillion rupiah in cash to Bank Central Asia overthe week following May 16 as deposits were with-drawn. On May 29, Bank Central Asia was broughtunder the auspices of IBRA, the owners’rights weresuspended, and an outside management team intro-duced. By the end of the month, runs on Bank Cen-tral Asia had decreased.

The specific nature of the attacks against BankCentral Asia was especially devastating to confi-dence in the banking sector, with many viewing therun on the bank as politically inspired.69 In this envi-ronment, other bankers sought to maximize their im-mediate liquidity to protect themselves in the eventof runs. The stock of vault cash increased, interme-diation declined even further, and interbank marketsbecame more segmented. With interest rates rising inthe face of uncertainty, banks bid up deposit rates tolevels substantially above those that they were ableto charge their borrowers. The sizable negative inter-est spread across much of the banking sector caused

60

68In the seventh bank, the state-owned Bank Exim, part of themanagement team was replaced. The bank now has become partof the new Bank Mandiri.

69Seventy percent of Bank Central Asia was owned by theSalim Group, the largest conglomerate in the country, and associ-ated with the Suharto government. Thirty percent of the bank wasowned by two children of the then President.

Appendix I

a continuing erosion of the capital base of the af-fected banks. Nevertheless, liquidity support fromBank Indonesia—apart from that to Bank CentralAsia—remained limited, and Bank Indonesia wasincreasingly successful in stabilizing monetary con-ditions in line with commitments under the IMF-supported program. By mid-July 1998, Bank In-donesia also reintroduced market-based monetarymanagement—with the start of auctions of one-month central bank bills—to enhance monetary con-trol, and thus helping to redistribute liquidity in thesegmented interbank market (three-month auctionsof central bank bills were started in October 1998).

Meanwhile, the bank restructuring process wasgiven a fresh impetus. A third head of IBRAhadbeen appointed, a finance ministry official with pre-vious experience in the banking sector. Continuity inthe restructuring process was assured with the ap-pointment of the first head of IBRAas finance min-ister in the new government. The new team neededto ascertain the true condition of the banks in orderto undertake appropriate remedial action. The banks’own reported figures were deeply unreliable, withmany banks still posting profits in early 1998 on thebasis of unrealized foreign exchange valuation gainsand lack of recognition of deterioration in their loanportfolios. In March 1998, Bank Indonesia had an-nounced new provisioning and classification guide-lines, broadly in line with international standards,but application was very patchy, in part because oflack of expertise both at Bank Indonesia and in thebanks.

International auditors were contracted—financedby the World Bank in the case of the 55 IBRAbanks,and by the Asian Development Bank in the case ofthe major non-IBRAbanks—to conduct portfolio re-views on international accounting standards byusing the new classification and provisioningrules.70As these were completed, they confirmed thepicture of deep and pervasive problems.

Toward a Comprehensive Resolution Strategy

Portfolio Review

Action toward the resolution of ailing banks re-sumed when audit results became available. In June1998, the results of the portfolio reviews of thecondition of the six private banks included in theseven largest borrowers of Bank Indonesia resources

(Box 14, Step 4, part A) showed that they had non-performing loans equivalent to at least 55 percent oftotal loans (over 90 percent in one large bank), andall the banks were proven to be deeply insolvent.71

For most of these banks the loan portfolios weredominated by connected lending. In July 1998, fourof the six banks were formally declared insolvent.On August 21, a resolution strategy was announcedfor these six banks. Three were “frozen” (Box 14,Step 6), while a rehabilitation program was definedfor the other three. One bank, Danamon, the secondlargest private bank in terms of number of depositorsbefore the crisis, was to be recapitalized and used asa “bridge” bank to receive deposits and assets fromsome of the closed banks. For the two remainingbanks, the former owners would be given a last op-portunity to clear their insolvencies; otherwise theywould be merged into Danamon or closed. Theprocess served also to highlight deficiencies in thelegal framework under which IBRAhad been oper-ating, with recalcitrant shareholders in the remainingtwo banks able to hold up the insolvency declara-tions in those banks. Ensuring proper powers forIBRA to assume control of insolvent banks was oneof the most important objectives of the amendmentsto the banking law prepared at the time.

By early August 1998, the results of the portfolioreviews for a group of 16 large banks, all of themnon-IBRA except for Bank Central Asia, were be-coming available. These showed that the bankswere, in general, clearly different from IBRAbanks,in terms of quality of risk controls, compliance withprudential norms, and overall quality of manage-ment. Nevertheless, the financial condition of thesebanks, too, was shown to be very weak, with manyof the banks insolvent. Given that many of thesebanks would have been expected to be among thestrongest in the country, these reviews confirmed thedeep insolvency of the banking system as a whole.

Legal and Supervisory Preparations

In October 1998, the parliament passed amend-ments to the banking law that modified previous re-quirements regarding bank secrecy and ended re-strictions on foreign ownership of banks. Theseamendments also enabled IBRAto operate effec-tively—for instance to be able to transfer assets andto foreclose against a nonperforming debtor.

Consistent with program commitments, substan-tial progress has also been achieved in reviewing

61

70Portfolio reviews on a similar basis were conducted by BankIndonesia supervisors themselves in the case of the smaller, non-foreign exchange banks.

71Results of these initial reviews were quickly leaked to thepress, causing consternation among the public. Security proce-dures, and other aspects of conducting the reviews, were tight-ened substantially in the light of experience with this first “wave.”

APPENDIX I

and strengthening the prudential and regulatoryframework on a number of critically important is-sues, as follows:

• Loan classification, loan provisioning, and thetreatment of debt restructuring operations.These three new regulations became effective asof the end of December 1998. Five loan cate-gories are defined, namely, pass, special men-tion, substandard, doubtful, and loss, with re-spective provisioning of 1 percent, 5 percent, 15percent, 50 percent, and 100 percent.

• Liquidity management. Banks are now requiredto submit a liquidity report twice monthly fortheir global consolidated operations. The reportcomprises a foreign currency liquidity profile,and a combined rupiah and foreign currencyprofile. The liquidity report collects data inweekly maturity bands for four weeks and forthe next two-month period. The report includesa section outlining the efforts the bank intendsto take to cover any liquidity shortfall or absorbany liquidity surplus.

• Foreign exchange exposure. The regulation hasbeen broadly satisfactory but there are certainshortcomings regarding implementation of theregulation. For example, Bank Indonesia may(rather than must) impose administrative sanc-tions on banks that do not comply with the lim-its for foreign exchange exposure.

• Connected lending.The newly published regu-lations tighten the rules in this field where themost widespread and damaging abuses tookplace in the period leading to the crisis, by andlarge in line with international best practices.

• Capital adequacy. The required capital ade-quacy level has been temporarily lowered to 4percent. The new regulation also ensures thatbanks can compute as supplementary capital(Tier 2) only general reserves for possible earn-ing asset losses up to a maximum of 1.25 per-cent of total risk-weighted assets, whereas theprevious regulation allowed banks to count asTier 2 both general and specific loan-loss provi-sion. The regulation will be reviewed by the endof 1999 to evaluate setting the date for banks tocomply again with a minimum capital adequacyrequirement of 8 percent.

• Disclosure of financial statements.Banks arenow required to publish their financial state-ments quarterly, beginning April 1999.

IBRA conducted forensic type audits to identifypossible irregularities and to examine the compli-ance with legal requirements of the 14 banks thenunder its control. On August 21, 1998, the govern-ment announced that former owners of 10 of thebanks that had been out of compliance with legalrequirements had one month to pay back the liquid-

ity support Bank Indonesia had provided, or besubject to further penalties. By late September,about 200 trillion rupiah of assets, at the owners’valuation, had been pledged from several of theseowners as well as about 1 trillion rupiah in cash.IBRA’s advisors valued the assets at 92.8 trillionrupiah, and IBRAtentatively announced full settle-ment with owners of three of the banks, includingBank Central Asia. These agreements, however,were not accepted by the government, whichsought a greater up front contribution of cash. Afterprotracted negotiations, the owners agreed to settletheir obligations within four years, but that assetsrepresenting these obligations would be transferredto IBRA to be placed into a holding company; itwas intended that 27 percent of the obligationswould be realized in the first year. With Bank Cen-tral Asia arriving at the first such agreement, otherformer owners reached similar agreements quickly.In mid-1999, however, owners of three banks weredeemed not to be negotiating seriously, and in fact,IBRA was preparing to take legal action againstthem.

Strategy for State Banks

In late August 1998 the authorities announced thatthe four state banks (including Bank Exim) underIBRA’s auspices would be merged into a single newbank, Bank Mandiri, under a management and oper-ational restructuring contract with a major interna-tional bank. The corporate business of a fifth, non-IBRA state bank, Bank Rakyat Indonesia (BRI), wasalso considered to be merged into this bank. BankMandiri would start with 30 percent of the assets ofthe banking sector. Although this share might de-cline as part of the restructuring, Mandiri would stillbe, by some margin, the largest bank in the country.

On September 30, 1998, Mandiri was establishedas holder of 100 percent of the shares of the fourcomponent banks. On February 12, 1999, a volun-tary severance scheme was offered to the headquar-ters staff of the four component banks as a prelude tointegrating their functions, and plans were finalizedfor centralizing two critical functions—treasury andcredit assessment. The bad loans from the compo-nent banks were transferred to the asset managementunit at the end of March 1999, and performing loanswere progressively transferred from the componentbanks into Bank Mandiri.

Meanwhile, the portfolio reviews confirmed thatthe three remaining state banks—Bank Negara In-donesia, BRI, and BTN—were also deeply insol-vent. The government announced plans to recapital-ize them, and in April 1999 presented a blueprint forthe restructurings. Bank Negara Indonesia is likelyto be the second largest bank in the country, with

62

Appendix I

about 15 percent of the sector. Meanwhile, the gov-ernment had also announced plans to recapitalize the27 regional development banks, which togethermade up 2 percent of the banking sector.

Bank Mandiri has made solid progress in imple-menting its restructuring plan, and is preparing to al-most halve its staff and close branches. Plans to en-hance loan recoveries have been prepared. A mergerof its four component banks is scheduled; capitaliza-tion of the bank is envisaged to be completed by theend of the year. Privatization is scheduled to beginwithin two years.

Meanwhile, the other state banks are in the processof finalizing blueprints for their restructuring. Re-capitalization is envisaged to take place betweenSeptember 1999 and March 2000. Privatization ofthe largest of these banks, Bank Negara Indonesia, isexpected in 2002. IBRAannounced the resolutionstrategy for the banks it has taken over. One smallbank will be merged into Bank Central Asia, andeight banks into Danamon. Bank Central Asia andthe “B” bank whose owners were unable to providetheir share of the recapitalization needs are expectedto be sold by the end of 1999. At that stage, there-fore, there will be only one IBRAbank (Danamon),which will serve as a “bridge bank” to wind downthe activities of the banks brought under its control.

Strategy for Private Banks

On September 30, 1998, on behalf of the govern-ment, Bank Indonesia announced a plan for thejoint recapitalization of those remaining privatebanks that met certain specified conditions. The ob-jective of this plan was to retain a residual privatebanking sector from among the “best” privatebanks, recognizing that given the economic turmoilthat has affected Indonesia, even well-run bankswould likely have run into serious difficulties.Moreover, the plan was designed to foster burdensharing between the private sector and the govern-ment regarding the cost of resolving these banks.The details are as follows:

• For those banks that had capital adequacy ratiosbetween negative 25 percent and 4 percent,business plans that demonstrated medium-termviability, and owners deemed to be “fit andproper,” the government indicated its willing-ness to contribute up to four rupiah for every ru-piah contributed by those owners to restore abank to having a capital adequacy ratio of 4 per-cent—the minimum capital adequacy require-ment for the end of December 1998. Owners’contributions would generally be in cash, whilethose of the government would be in bonds.

• The government announced that it would obtainits equity stake through preference shares that

would be convertible into ordinary shares in ei-ther of two situations—if the bank failed tocomply with the targets of its own businessplan, or after a period of three years. During thethree-year period, the owners would be able toreacquire their shares in the bank by repayingthe government’s contribution (in effect, thecontribution would have been a loan), either fortheir own account or for an outside investor. Atthe end of the three years, the governmentwould seek an independent valuation of thebank, and the owners would have the first op-tion to buy back the government’s shares. Oth-erwise, the government would sell its shareswithin a specified period. To encourage ownersto put in new capital, the government pledged toallow owners to retain day-to-day control ofbanks.

• In addition, once the joint recapitalization wasagreed upon, all category 5 loans (i.e., thoseloans classified as “loss”) were to be transferredat a zero price to the asset management unit,which would enter into a contract for the recov-ery of the loans with the originating bank. Attheir discretion, the banks could also transfercategory 4 loans (those classified as “doubtful”)at zero price to the asset management unit, forhandling on the same basis as the category 5loans. Any recoveries from such loans would beused immediately to buy back the government’spreference shares, thus giving the governmentthe prospect of an early return of its financial in-fusion, and reducing the amount to be paid bythe owners to reacquire full control of theirbank.

In December 1998, the plan was threatened to bereversed when the President announced instead astrategy of forced mergers. With the uncertainty inthe markets, as well as the worst rioting since May1998, leading to renewed depreciation of the rupiah(from about 7,500 rupiah to the U.S. dollar to 9,000rupiah per U.S. dollar), the private bank recapitaliza-tion plan was finally reaffirmed. In February 1999,Parliament passed a budget that included 34 trillionrupiah for bank restructuring (17 trillion rupiah netof expected recoveries).

By that time, work on assessing business plans—the fitness and propriety of owners and managersand availability of capital infusions—had beencompleted, and bank closures were widely regardedas imminent. Indeed, ministers had made it clearthat banks that failed the tests for the recapitaliza-tion plan would be closed on February 26, 1999.However, intense lobbying from owners of banksthat realized that they were in danger of beingclosed succeeded in postponing the government-scheduled interventions at the last minute. The pub-

63

APPENDIX I

lic recognized the cause of the postponement, andamid clear indications of government disarray therupiah fell another 10 percent, to almost 10,000 perthe U.S.dollar.

On March 13, 1999, the government announcedthe closure of 38 banks (comprising 5 percent of thebanking sector) and the takeover by IBRAof sevenothers (comprising 2 percent of the sector). Ninebanks (or 10 percent of the banking sector) weredeemed eligible for recapitalization under approvedbusiness contracts. The banks subsequently pro-vided their share of capital by April 21. At the time,73 banks (5 percent of the banking sector) had capi-tal adequacy ratios of at least 4 percent and, hence,did not need government support. Although theterms of the original recapitalization plan were thusbroadly reaffirmed, the government did not explic-itly restate its earlier commitment to leave the day-to-day running of the banks in the hands of theowners.

The closures and takeovers were handled effi -ciently by Bank Indonesia and IBRA, although theagencies had difficulty transferring the deposits todesignated recipient banks over the weekend ofMarch 13, since staff in a number of banks deniedthem access in order to increase their bargainingpower for higher severance payments. IBRAunder-took to pay twice the legal minimum, and indicated

that severed staff should negotiate with the formerowners for anything more. Although over the fol-lowing weeks there were public disturbances by em-ployees of the closed banks, Bank Indonesia/IBRAgradually achieved access to the remaining branchesof the banks.

The March 13, 1999 announcements indicatedalso that the “A” category banks would be subject toreview regarding the fitness and propriety of theirowners and managers and, for some of them, thesources of their capital infusions. The sources ofcapital were verified by April 21. However, one-third of the owners, managers, and commissionersfailed the “fit and proper” test. These commissionersand managers were required to be replaced immedi-ately and failed owners themselves were given up to90 days to divest their holdings.

Of the nine banks eligible for recapitalizationunder the government’s plan, owners of seven bankssupplied their share of the necessary additional capi-tal in advance of the April 20 deadline. IBRAbe-came involved in intense discussions as the deadlineapproached to resolve the situation of the remainingtwo banks. Ultimately, one of these banks (holdingabout 1 percent of total bank assets) was acquired bya major foreign bank, the largest such acquisitionsince the banking crisis began; for the other—dis-cussions on which were hampered by the absence of

64

44 Joint ventures andforeign banks (8%)

34 Public domesticbanks (42%)

160 Private domesticbanks (50%)

Commercial banks at the end of June 1997(share of liabilities)

40 Joint ventures andforeign banks (10%)

43 Public domesticbanks (73%)

82 Private domesticbanks (17%)

Commercial banks at the end of July 1999(share of liabilities)

Figure 9. Indonesia: Progress in Financial Sector Restructuring

Source: National authorities; and IMF staff estimates.Note: As of the end of July, 60 private and three public domestic banks and three joint venture and foreign banks had been closed.The state had in-

tervened in 14 private and 15 public domestic banks and in one joint venture/foreign bank. Four private domestic banks had exited through merger.Categories are not mutually exclusive.

Appendix I

a controlling shareholder—no sources of capitalcould be found in time, and the bank was taken overby IBRA.

Over the following month, IBRAnegotiated per-formance contracts and memoranda of understand-ing with the managements of the eight remainingbanks. The government had decided that it wished tohold its shares in these banks as ordinary stock,rather than convertible preference shares, but thesememoranda of understanding provided essentiallythe same protection to the owners of the banks—thatis, the owners retained the day-to-day control of thebanks, and various areas of the government retainedthe right to become involved if necessary. Mean-while, updated audits provided revised figures forthe banks’insolvency, indicating capital needs al-most double the amount earlier estimated.

At the end of May 1999, the owners of the banksprovided their share of the extra capital require-ment, and the government issued bonds to provideits share of the necessary recapitalization.72 In thecase of the listed banks, a rights issue underwrittenof up to 80 percent by the government served to re-duce need for public funds; for the largest recapital-ized bank, this issue was particularly successful inattracting new private investment. The governmentnow owns between 60 and 80 percent of the stockof these banks; at the end of three years indepen-dent valuations of these banks will be conducted,and the owners will be given the right of first re-fusal to buy back the government’s shares. If theowners decide not to purchase the shares, the gov-

ernment will sell its shares over the following year.In the meantime, the government can only dilute itsshareholding with the concurrence of the owners.The government’s share will also decline during thethree years, representing the loan losses realizedthat will be subsequently transferred to IBRA. Dur-ing this period, the owners may themselves buyback the government’s shares, at the price the gov-ernment paid plus interest. With the prices of thesebanks shares now rising strongly, it appears that theowners of at least one of the banks are alreadyseeking to buy back the government’s shares. Thus,the recapitalization program has saved these banks,and the government seems likely to get back atleast a share of its investment sooner than origi-nally envisaged.

State of Restructuring Efforts

The Indonesian banking sector has been consoli-dated significantly in the last two years (Figure 9).Since mid-1997, the number of private domesticbanks has been nearly halved through closures orstate takeovers. Reflecting this consolidation, banksunder state control now hold about 70 percent of lia-bilities, compared to 40 percent before the crisis.

Public Cost of Financial Sector Restructuring

As of mid-1999, the public contribution to finan-cial sector restructuring has been equal to 51 percentof GDP(Table 15). The largest share of this has beenused to recapitalize banks and provide liquidity sup-port. The expected budgetary cost of bank restructur-ing (i.e., the interest cost on the government bonds)in the 1999/2000 fiscal year is 34 trillion rupiah or3 percent of GDP; half of this is to come from recov-

65

72Except for the bank being taken over by the foreign bank,where the capital infusion will follow once the foreign bank hascompleted its due diligence.

Table 15. Indonesia: Public Cost for Financial Sector Restructuring(as of mid-1999)

Percent of GDP In Billions of U.S. dollars1

Central bank liquidity support assumed by the budget 12 20

Recapitalization including outlays for deposit guarantee 23 40

Purchases of nonperforming loans or capital for asset management company 12 20

Interest cost (on budget) 3 5

Total 51 85

Source: National authorities; and IMF staff estimates.1Converted at 7,500 rupiah per U.S. dollar.

APPENDIX I

eries, mainly from shareholder settlements. IBRAhas been investigating all failed banks (those closedand those taken over) to see if there were violationsof insider lending limits, and, if so, will pursue theformer owners of these banks to provide assets ascompensation for these violations. Agreements havealready been reached with the former shareholders ofeight such banks, and negotiations with several moreare in progress. Two of these agreements have so farresulted in assets being transferred from two of thesebanks to holding companies under the joint controlof IBRA and the former owners. Sales from these as-sets are likely to begin in the coming months, withover 12 trillion rupiah expected by March 2000.

Meanwhile the state and IBRAhave intensifiedtheir loan recovery efforts. Each institution has pub-lished the names of their largest debtors and invitedthem to begin negotiations. IBRAaimed at classify-ing all debtors into four categories by the end of Au-gust 1999, depending on their degree of cooperationand business viability. IBRA hoped to adopt a strat-egy for each debtor based upon its classification.Noncooperating debtors will be widely publicized.An interdepartmental committee has been estab-lished to coordinate the preparation of a database onthe debtors, and to facilitate the establishment of ajoint-creditor negotiating position, with the debtorsunder a lead creditor institution.

66

Korea, a largely industrialized country and mem-ber of the Organization for Economic Coopera-

tion and Development (OECD), was initiallythought immune from the financial sector problemsin its less developed neighboring countries.73 Never-theless, by late 1997, structural similarities and theforces of contagion contributed to the extension ofthe crisis to the Korean economy.

Background

Korea’s exchange rate remained broadly stablethrough October 1997. However, the high level ofshort-term debt and the low level of usable interna-tional reserves made the economy increasingly vul-nerable to shifts in market sentiment. While macro-economic fundamentals continued to be favorable,the growing awareness of problems in the financialsector and in industrial groups (chaebols) increas-ingly led to the difficulties for the banks in rollingover their short-term borrowing. Declining rolloverrates brought this ratio down to less than half at theend of 1997. International reserves decreased fromthe equivalent of 2.6 months of imports in 1993 totwo months in 1996, and—with the onset of the cri-sis—dropped to 0.6 months in 1997. Korea widenedthe exchange rate band on November 17, 1997; thewon fell sharply, and the Bank of Korea had to pro-vide large amounts of foreign exchange for Koreanbanks to honor their overseas commitments.

On December 4, 1997, Korea entered into a three-year Stand-By Arrangement with the IMF, amount-ing to $21 billion, augmented by arrangements withthe World Bank and the Asian Development Bank.Several countries pledged additional $22 billion as asecond line of defense. Agreement on the program—which also included far-reaching steps toward finan-cial sector rehabilitation—initially failed to increaserollover rates of short-term foreign debt, and thewon fell further. However, the combination of an

agreement with private foreign bank creditors on avoluntary rescheduling of short-term debt (con-cluded by the end of January 1998) and a rephasingof the IMF arrangement to allow an advancement ofdrawings succeeded in alleviating short-term foreignexchange pressures and permitted stabilization tobegin. Progress in stabilization allowed a gradual re-duction in interest rates, now standing at below pre-crisis levels. The won, which had depreciated closeto 2,000 won per U.S. dollar, has stabilized in therange of 1,200 won per U.S. dollar. Useable reserveshave increased to $52 billion as of mid-1999.Growth is now expected to be about 9 percent in1999.


Prior to the crisis, Korea’s macroeconomic perfor-mance was generally praised, with GDPgrowing atan average rate of 8 percent in 1994–97. An increasein investment and exports fueled the growth. Unem-ployment remained low, averaging slightly above 2percent during the period, and inflation was stable atabout 5 percent. The fiscal position appeared to bestrong, and public debt was below 11 percent ofGDP, of which only about one-fifth was foreigndebt. Developments in the external sector followedclosely the evolution of the yen, which in 1995–96led to a widening of the current account deficit to theequivalent of almost 5 percent of GDPin 1996 andwhich declined the next year to 2 percent. This ex-ternal deficit was financed by private capital in-flows. The ratio of total external debt to GDPin-creased significantly from 20 percent of GDPin1993 to 33 percent in 1996 and to 35 percent in1997. At the same time, the proportion of short-termdebt to total debt increased, amounting to about two-thirds of total debt in 1996.


The Korean financial system was unusual amongemerging markets for its diversity. It comprised pri-vate commercial banks, a number of government-owned specialized and development banks, and a

Appendix II Korea

67

73This appendix draws on Baliño and Ubide (1999) and on con-tributions from Peter Hayward and Leslie Teo.

APPENDIX II

wide variety of very sizable nonbank financial insti-tutions. These institutions were closely interlinked.In particular, commercial banks had significant off-balance sheet exposures to nonbank financial institu-tions through their holdings or guarantees of com-mercial paper and corporate bonds underwritten bymerchant banks.

Commercial banks accounted for just over half theassets of the financial system. As of September1997, the sector comprised 16 nationwide banks, 10regional banks, and 52 foreign banks. The top eightbanks accounted for about two-thirds of commercialbank assets (excluding trust accounts). Regionalbanks were established to develop specific regions,particularly to foster the growth of small- andmedium-sized enterprises. Foreign banks have beenallowed to open branches since 1967, although theirmarket share remained very small. Commercialbanks were supervised by the Office of Banking Su-pervision at the Bank of Korea.74 Commercial banksalso operated trust accounts that were separately ac-counted for but were managed like the rest of theirbanking business.75 Trust accounts had grownrapidly in recent years (they accounted for close to40 percent of total commercial bank assets as of theend of 1997), largely because they had been less reg-ulated and had been able to offer higher interest ratesthan regular commercial bank business.

Specialized and development banks were estab-lished in the 1950s and 1960s to provide funds tospecific strategic sectors.76 They accounted forabout 17 percent of financial system assets. Al -though specialized banks could borrow from thegovernment, deposits constituted their main sourceof funding. Funding for development banks, whichare wholly government-owned, came mainly frombonds issued domestically and abroad.77 The KoreanDevelopment Bank, the largest development bank,was established in 1954 to supply long-term credit tomajor industries. Its assets were heavily concen-trated in lending to large corporations, mainly fi-

nancing fixed investment (including infrastructureprojects). These banks were traditionally not subjectto the same prudential standards and supervision ascommercial banks, and were overseen by the min-istry of finance and economy.

Meanwhile, nonbank financial institutions com-prised 30 percent of financial system assets at theend of 1997 and consisted of three types of institu-tions: investment companies, savings institutions,and insurance companies. Of these, investment in-stitutions, which consisted of merchant banks, in-vestment trust companies, and securities compa-nies, were the largest in terms of assets, followedclosely by savings institutions. Many nonbank fi-nancial institutions were predominantly owned, di-rectly or indirectly, by chaebols and other largeshareholders. They were used to finance activitieswithin the chaebol group, and became an increas-ingly important source for intermediating chaebolnotes and other paper. Nonbank financial institu-tions and trust accounts at commercial banks pro-vided a means to circumvent restrictions on com-mercial bank intermediation.

Precrisis Weaknesses in the Financial System


During the 1960s and 1970s, Korea embarked onan outward-oriented industrialization strategy spear-headed by the chaebols, which were supported bypreferential access to subsidized credit (so-calledpolicy loans).78 Interest rates were administered, fi-nancial innovation was restricted, and competition inthe banking system was limited. These governmentpolicies resulted in a tightly controlled, government-administered financial system characterized bychronic excess demand for credit.

Since the mid-1970s the government took steps—including the granting of tax preferences—to de-velop capital markets and thus to reduce corporatereliance on bank borrowing. A program of gradualdomestic financial sector reform was introduced inthe early 1980s, including bank privatization andderegulation of banks’activities, abolition of creditcontrols, introduction of new financial instruments,granting of greater business opportunities to the non-bank financial institutions, and partial interest rate

68

74The banks’trust account business and leasing affiliates werenot under the supervision of the Office of Banking Supervision.

75In 1970, the trust business was assigned exclusively to KoreaTrust Bank, which merged with Seoul Bank in 1976. By the endof 1995, all deposit money banks and development institutions,with the exception of the Korea Export-Import Bank (KEXIM)and foreign banks, were allowed to engage in trust businesses.

76There are four specialized banks—the Industrial Bank ofKorea and three other banks centered on agricultural, fisheries,and livestock cooperatives. Development banks comprise the Ko-rean Development Bank, KEXIM, and the Long-Term CreditBank.

77Although those bonds did not carry a formal governmentguarantee, the government was legally obliged to ensure that thebanks were always in a position to meet their liabilities when theyfell due.

78The National Investment Fund was created in 1974 for thispurpose, and was funded by the compulsory deposit of savingsfrom pensions, savings and postal savings accounts, and byother purchasers of National Investment Fund bonds, such as lifeinsurers.

Appendix II

deregulation. At the same time, many controls re-mained, particularly on commercial bank lendingand interest rates. In an effort to reduce the relianceof thechaebols on bank borrowing, the governmenttightened the credit control system by setting a ceil-ing on the share of bank credit to chaebols. In addi-tion, banks were required to meet minimum credittargets (initially set at one-third of new lending) forsmall- and medium-sized enterprises.

Beginning in the early 1980s, government in-volvement in bank lending decisions was graduallyreduced, but banks developed few skills in creditanalysis or risk management. Lending decisionswere still largely based on the availability of collat-eral rather than on an assessment of risk and futurerepayment capacity. Because of their large expo-sures and inadequate capitalization, banks were gen-erally in a weak position relative to their chaebolclients. Reflecting the history of directed lending,banks did not insist on, or receive, full financial in-formation from chaebols. In addition, basic account-ing, auditing, and disclosure practices were signifi-cantly below international best practice (forexample, consolidated accounting or marking tomarket were mostly absent in Korea). Furthermore,the strong role of banks in the bond market, alongwith the bonds’relative illiquidity and the lack oftransparency in the equity market (due to lax disclo-sure standards), impeded the capital markets’role inensuring sound corporate governance. Meanwhile,the banks were rapidly expanding their foreign oper-ations and becoming subject to significant liquidityrisk; failure to manage this risk was a probable causeof the crisis.

The high leverage ratios of thechaebols and theirlow profitability made them very vulnerable to anyshock to their cash flow.79 The health of the bankingsystem, in turn, was extremely dependent on the via-bility of the chaebols. Banks were highly exposed tothem, both directly through loans and discounts, andindirectly through the guarantee of corporate bondsand commercial paper.80

A large portion of the foreign borrowing bybanks, particularly by merchant banks, was under-taken through overseas subsidiaries and foreign

branches.81 Several factors explain the reliance onshort-term capital inflows:

• The capital account had been only partially lib-eralized, with intermediation through domesticbanks favored over foreign direct investmentand direct corporate borrowing.82 Restrictionsagainst short-term foreign borrowing by finan-cial institutions were relaxed, while limits onlong-term borrowing and foreign participationin domestic equity and bond markets were re-tained, encouraging the development of largematurity mismatches in banks’balance sheets.At the end of December 1997 short-term assetscovered only about half of short-term liabilitiesin commercial banks, and 25 percent in mer-chant banks, despite the introduction of a re-quired minimum ratio of 70 percent.

• Increased access to trade credits and deregula-tion permitted the use of trade credits for work-ing capital. There was a seven-fold increase intrade credits during 1994–96, only partly ac-counted for by the rapid growth in tradevolume.

• The substantial nominal interest differential infavor of dollar and yen borrowing was rein-forced by the expectation of a stable exchangerate resulting from the won’s managed peg tothe dollar. In addition, the short-term risk pre-mium was lower than for longer maturities, andshort-term funds could be raised relatively eas-ily in international money markets.83 Thus, do-mestic banks relied on external short-term fundsto finance long-term domestic investments.

Weaknesses in Prudential Regulation and Supervision

Prudential regulations and, especially, supervisionwere not strengthened when the banks were grantedgreater independence in lending decisions and whendomestic financial markets and the capital account

69

79The debt ratio of most chaebols exceeded 400 percent duringthe 1990s, compared with an average of 210 percent in Japan, 150percent in the United States, and 90 percent in Taiwan, Provinceof China. Low profits decreased their ability to service this debt;operating cashflow as a percentage of interest payments was only80 percent in 1996.

80At the end of 1997, the 30 largest chaebols accounted forhalf, and the five largest chaebols for one-third, of the corporatedebt outstanding. The top 30 chaebols were responsible for about30 percent and the top five chaebols for about 18 percent of com-mercial bank loans at the end of 1997.

81Foreign liabilities of domestic financial institutions, includ-ing those of overseas subsidiaries and foreign branches, increasedfrom $40 billion at the end of 1993 to $160 billion by the end ofSeptember 1997.

82There were restrictions on issuance of securities abroad—only corporations that had obtained an international credit ratingof BBB and above could undertake such issuance—and on con-tracting loans at spreads higher than 100 basis points overLIBOR. These regulations strengthened the role of the major Ko-rean banks (whose ratings benefited from implicit governmentsupport) as the conduits of external finance to domestic corpora-tions.

83International interbank lending to Korea of about $15 billionin 1994 and about $25 billion in 1995–96, to reach a total of$108.5 billion at the end of 1996. Of this amount, about 70 per-cent had a maturity of less than a year.

APPENDIX II

were liberalized. A reform plan included several pro-posals to bring prudential regulations closer to inter-national standards. Commercial banks were requiredto reach a minimum of 8 percent capital adequacyratio by the end of 1995. In 1995, the Office ofBanking Supervision introduced a reporting systembased on the CAMELframework, designed to giveearly warning of problems. In addition, the govern-ment introduced, effective January 1997, a depositinsurance scheme funded by the financial institu-tions. The scheme provided full coverage for all in-sured deposits, up to 20 million won per individualdepositor. Despite these reforms, several aspects ofthe Korean supervisory and regulatory frameworkdiminished the effect of these improvements.

Supervision was fragmented.Commercial bankswere under the direct authority of the monetary board(the governing body of the Bank of Korea) and theOffice of Banking Supervision. However, specializedbanks and nonbank financial institutions were underthe authority of the ministry of finance and economy,although the ministry delegated on-site examinationof some nonbank financial institutions to the Officeof Banking Supervision. This lack of a unified sys-tem of supervision and regulation, together with theweak supervision performed by the ministry on non-bank financial institutions, created conditions favor-able to regulatory arbitrage and high-risk practices,especially among commercial banks’trust businessand merchant banks. In addition, the supervisory au-thority had the power to waive requirements, whichnot only facilitated forbearance but also made en-forcement nontransparent.84

Standards for loan classification and loan-lossprovisioning were rather lax. Nonperforming loanswere defined as loans that had been in arrears for sixmonths or more. Published data on bad loans onlyincluded those nonperforming loans not covered bycollateral.85 The classification system was based onthe loan’s servicing record and the availability ofcollateral without regard to the borrower’s future ca-pacity to repay. Provisions were based on credit clas-sification and consisted of 0.5 percent of “normal”credits, 1 percent of “precautionary” credits, 20 per-cent of “substandard” credits, and 100 percent of“doubtful” and estimated “loss” credits. However,the minimum provisioning requirements were to bephased in over a number of years ending at the endof 1998. Losses were not expected to exceed 2 per-

cent of total loans, and any excess provisioning over2 percent was not tax deductible. As a result of theseregulations, aggregate provisions were still under 2percent at the end of 1997.

Provisioning rules and accounting standards forsecurities holdings also fell short of internationalbest practice. Banks’books recorded securities atcost: mark-to-market accounting was not fully ap-plied. Meanwhile, the banks had large unrecordedlosses arising from their equity holdings during a pe-riod of declining market prices. Furthermore, therewas no consolidation of statements encompassing aparent bank and its subsidiaries.

The lax limitations on risk concentration facili-tated the highly leveraged corporate finance struc-ture of Korean conglomerates.A 1991 revision ofthe General Banking Act set the limit for single bor-rowers at 20 percent of a bank’s capital for loans and40 percent for guarantees, with a very generousgrandfathering clause and a phase-in period of threeyears. The grandfathering arrangements were ex-tended in 1994 and 1997. Limits on lending to bigconglomerates were set bank-by-bank under the“basket control system,” under which the shares ofloans to the top 5 and 30 business groups relative tototal loans of the bank should not exceed the ratiosset by the Office of Banking Supervision. These lim-its were tightened in August 1997; they limited lend-ing to a single chaebol (including guarantees) to 45percent of banks’capital for commercial banks and150 percent for merchant banks. Not only were thesetighter regulations still lax in comparison to those inother OECD countries but many banks continued tobreach them.

Restoring the Soundness of theFinancial Sector

Emergency Measures

From the late summer of 1997, international cred-itors began to reduce their exposure to Korean finan-cial institutions and to withdraw their short-termcredit lines. The devaluation of the Thai baht in July1997, the subsequent contagion to other regionalcurrencies, and the crash of the Hong Kong stockmarket in late October 1997 sent shock waves to theKorean financial system. Market confidencedropped sharply and rating agencies downgradedKorea’s sovereign status.

In August 1997, the Korean authorities announcedthat they would ensure that Korean financial institu-tions would meet their foreign liabilities. Nonethe-less, the withdrawal of foreign credit lines intensi-fied in the ensuing weeks. Faced with increasingdifficulties in meeting their short-term foreign oblig-

70

84Supervisors sometimes allowed only partial application ofregulations, such as provisioning requirements, to avoid weaken-ing banks’earning reports.

85Nonperforming loans, as a percentage of total loans, wereonly 0.8 percent in 1996; however, if “substandard” loans wereincluded, the figure, even using Korean definitions, would havebeen 4.1 percent.

Appendix II

ations, banks turned to the Bank of Korea for foreignexchange liquidity support.

Meanwhile, the government had deliberatelystepped back from direct intervention and did notbail out failing chaebols. From the beginning of1997, an unprecedented number of the highly lever-aged chaebols went into bankruptcy, dragged downby excessive investment, declining profits, and asubstantial debt burden. By the end of November,six of the top 30 chaebols had filed for court protec-tion; a seventh went into bankruptcy in December.These large bankruptcies, together with rising bank-ruptcies among small- and medium-sized enter-prises, further damaged the asset quality of financialinstitutions.

By the fall of 1997, the balance sheets of Koreanfinancial institutions had deteriorated severely. Theshare of nonperforming loans in total assets of com-mercial banks had increased by about 70 percent be-tween December 1996 and September 1997 andamounted to about 80 percent of banks’capital.86 Asa result, the net worth of many financial institutionsfell perilously low, and a significant shortfall in cap-ital adequacy emerged.87 Of the 26 commercialbanks, 14 had capital adequacy requirements below8 percent, of which two were deemed to be techni-cally insolvent (with zero or negative capital). In ad-dition, 28 of the 30 merchant banks had capital ade-quacy requirements below 8 percent and 12 weredeemed technically insolvent.

During November and December 1997, the Bankof Korea placed some $23 billion of official reservesin deposit at foreign branches and subsidiaries of do-mestic financial institutions, which the banks used torepay their short-term debt that they could notrollover. The Bank of Korea’s usable official re-serves were thus quickly depleted. On November 19,the government attempted to calm markets by an-nouncing a reform package that included a wideningof the daily exchange rate band to +/– 10 percent(from +/– 2!/4 percent) and measures to purchasenonperforming loans. Market concerns remained,

however, and during the last week of November thedepletion of international reserves intensified tosome $1 billion to $2 billion a day, driving usable re-serves to only $5 billion by the end of the month.

To maintain public confidence in the banking sys-tem, in mid-November 1997, the government guar-anteed all deposits of financial institutions until theend of 2000, and announced that it would providetemporary liquidity support to banks when needed.Strengthening the financial sector was a key compo-nent of the IMF-supported program adopted in De-cember 1997.

Institutional Changes

The government also completed important re-forms of the institutional arrangements, which hadbeen recommended by the Presidential Commissionon Financial Reform earlier in 1997. Many of thesereforms became part of the IMF-supported program.These reforms included:

• Laws passed in December 1997 significantlystrengthened the independence of the Bank ofKorea; consolidated all financial sector supervi-sion (for commercial banks, merchant banks, in-surance companies, securities firms, and othernonbank financial institutions) in a single Fi-nancial Supervisory Commission separate fromthe government; and merged all deposit insur-ance protection agencies into the Korea DepositInsurance Corporation (KDIC), a new agency.The new supervisory agency was established intwo stages to give time for the necessary prepa-ration and not to detract unnecessarily from themanagement of the crisis. The Commission it-self was established in April 1998, and the vari-ous supervisory bodies fell under its control asof that date. However, the full unification of su-pervisors, as the Financial Supervisory Service(FSS), with concomitant and extensive manage-ment and structural changes, took place only asof January 1, 1999. In April 1999, the opera-tional autonomy of the Financial SupervisoryCommission was strengthened with the passageof legislation to grant it the power to license andde-license financial institutions, as well as to su-pervise specialized and development banks.

• In early 1998, the government established a Fi-nancial Restructuring Unit within the FinancialSupervisory Commission to oversee and coordi-nate the restructuring of the financial sector. Asimilar unit was also set up to spearhead thegovernment’s efforts to restructure the financialposition of the weaker chaebols. The KDIC wasprovided with powers and funds to pay back de-posits in failed institutions and, if necessary, toalso provide recapitalization funds to banks.

71

86Nonperforming loans are a lagging indicator of the soundnessof the banking sector, especially when loans are only classified asnonperforming after having been in arrears six months, ratherthan the usual three months. A more valid measure is the ratio ofdishonored bills and checks, which more than doubled during thesame period and increased fivefold in the last quarter of 1997.

87One measure of balance sheet deterioration is the shortfall incapital adequacy represented by the amount of funding needed tobring a bank’s ratio of capital to risk-weighted assets to the mini-mum of 8 percent recommended by the Basel Committee forBanking Supervision. Estimates based on the end of September1997 balance sheet data showed, under Korean provisioning andloan classification rules, a shortfall of some 11.3 trillion won (3.0percent of 1997 GDP) for commercial banks, merchant banks, de-velopment, and specialized banks.

APPENDIX II

• A bridge bank (Hanaerum Merchant Bank) wascreated at the end of December 1997 to takeover the assets and liabilities of closed merchantbanks. The role of the Korean Asset Manage-ment Corporation (KAMCO) was expanded toenable it to purchase impaired assets from all fi-nancial institutions (Box 15).

A Bank Restructuring Plan

As part of the financial sector restructuring pro-gram, the authorities examined financial institutionsto determine their solvency; several groups of insti-tutions were also subject to external diagnostic re-views. Based on these examinations, the FinancialSupervisory Commission required weak institutions

to submit rehabilitation plans that had to be ap-proved by evaluation committees set up by the Fi-nancial Supervisory Commission. These plans speci-fied how the affected institutions would raise capitaland operationally restructure to improve profitabil-ity. Institutions whose plans were approved signedmemoranda of understanding with the Financial Su-pervisory Commission promising to meet targets setforth in their rehabilitation plans. For institutionswhose plans were not approved, the Financial Super-visory Commission developed and implementedseveral exit strategies.

The government undertook to commit public re-sources for bank recapitalization only under limitedcircumstances. After June 30, 1998, public re-sources—through subscription of capital instru-

72

The Korean Asset Management Corporation(KAMCO) was established in 1962 to collect nonper-forming loans for banks. In November 1997, legislationwas passed to create a new fund under KAMCO, sup-ported by contributions from financial institutions andgovernment guaranteed bond issues. This KAMCO-ad-ministered fund was given the mandate to purchase im-paired loans from all financial institutions covered by adeposit guarantee.

On August 10, 1998, a major reorganization ofKAMCO as a “bad bank” was completed with a viewto strengthening its asset management and dispositioncapabilities. KAMCO adopted a structure similar to theU.S. Resolution Trust Company, providing additionalbusiness functions such as workout programs for non-performing loans and more efficient asset disposal. Toenhance the transparency and the efficiency of its oper-ations, KAMCO has its accounts audited semiannuallyand publishes the results.

As of mid-June 1999, KAMCO has purchased assetswith a face value of 46 trillion won (11 percent ofGDP). Its purchases comprise two categories of assets:(1) “general” assets of companies currently operatingand (2) “special” assets, which correspond to cases thatare currently in court receivership and account for 70percent of the total portfolio. Only 20 percent of thesespecial assets have been finally resolved by the courts.Until recently, KAMCO only purchased won-denomi-nated assets, owing to lack of funding capacity in for-eign exchange. To overcome this deficiency, KAMCOfor the first time issued U.S. dollar denominated bondsin late December 1998 for $513 million to purchaseforeign currency denominated assets from commercialbanks.

For the purchase of nonperforming loans, KAMCOpays 45 percent of the book value of the underlyingcollateral, which is the average price obtained in auc-tions of similar collateral in the market. For unsecured

loans, the price is set at 3 percent. The prices for the or-dinary nonperforming loans are final. However, mostof the loans purchased so far have been special loansand for these types of loans, KAMCO pays 45 percentof the face value of the loans; but once a court orderedrepayment schedule is implemented, the price of thepurchase is readjusted to reflect the present value of thesettlement.

KAMCO’s sale strategy is to dispose of nonperform-ing loans in the fastest way possible, but in a mannerthat maximizes recovery value. KAMCO has used fourmethods to collect on its assets: it has sold nonperform-ing loans to international investors, foreclosed and soldunderlying collateral; sold nonperforming loans in apublic auction; and collected on loans. As of June 1999,KAMCO had recovered—through sales and collec-tions—about 9 trillion won from loans with a facevalue of 17 trillion won. Details on these operations areprovided in the table below.

Box 15. Korea: KAMCO Operations

Disposition and Sale of KAMCO’s Assets(In trillions of won, as of June 1999)

Price AmountFace Paid by Recovered

Disposition/Sale Amount KAMCO by KAMCO

International sale 2.9 1.1 1.2Foreclosure auction 2.1 1.0 1.0Public sale 0.2 0.1 0.1Collection 11.3 5.9 6.5Total 16.5 8.1 8.8

Source: KAMCO

Appendix II

ments and nonperforming loan purchases—couldonly be committed in the context of approved recap-italization plans, and on the condition that adequatecontributions be made by shareholders and otherstakeholders (exceptions would be made only underwell-specified conditions).

The program was implemented in stages, startingwith the most serious problems—the clearly insol-vent merchant banks. Simultaneously, the govern-ment announced a timetable for merchant banks andcommercial banks to be evaluated, and to attain mini-mum capital adequacy requirements. From mid-1998,steps were taken to restructure the remaining non-bank financial institutions, in particular, investmenttrust companies and life insurance companies. Mea-sures were adopted to strengthen prudential regula-tions and supervision, particularly in the areas of loanclassification and provisioning, foreign exchange liq-uidity, large exposures, and connected lending.

Merchant banks

In mid-December 1997, at the height of the crisis,the government announced the suspension of 14merchant banks, of which 10 were closed the follow-ing January. The bridge bank, Hanaerum MerchantBank, took over and liquidated their assets.88 The re-maining 20 merchant banks were required to submitrehabilitation plans that demonstrated the ability togradually strengthen their capital adequacy require-ments. On the basis of these plans, four merchantbanks were closed by the end of April, 1998. The re-maining 16 merchant banks were required to meetcapital adequacy requirements of 6 percent by theend of June 1998 and 8 percent by the end of June1999. In July and August 1998, the Financial Super-visory Commission conducted examinations to en-sure that merchant banks were complying with theirplans. As a result of these examinations, two morebanks were closed. Subsequently, two merchantbanks announced mergers with commercial banksand one more was closed in June 1999, leaving atotal of 11 merchant banks.

The government has not directly committed re-sources to recapitalize merchant banks in view of

their small size and the fact that many are owned bychaebols. Rather, these remaining merchant bankshave raised significant amounts of capital from cur-rent owners between the end of December 1997 andthe end of June 1999, with capital increasing from0.4 trillion won to 2.5 trillion won.

Commercial banks

In December 1997, the government took over twolarge commercial banks, Korea First Bank and SeoulBank, which were technically insolvent. Given theirsystemic importance, the government recapitalizedthem, and following the approval of requisite legis-lation, wrote down the equity of existing sharehold-ers by a factor of about 8:1 and removed managersresponsible for the losses. The government and theKDIC injected capital, acquiring a stake of about 94percent in each bank. Since March 1998 the bankshave been prepared for privatization with the help offoreign advisors; memoranda of understanding tosell them have been signed with Newbridge Capitaland HSBC (in late 1998 and early 1999), but thedeals have yet to be finalized (Box 16).

In early 1998, 12 commercial banks that did notmeet the minimum capital adequacy requirement of8 percent at the end of 1997 were required to submitrecapitalization plans, which the Financial Supervi-sory Commission evaluated with the help of interna-tionally recognized accounting firms. On June 29,1998 the Financial Supervisory Commission an-nounced the decisions on the recapitalization plans.Five small- to medium-sized banks were closed,with their assets and liabilities transferred to fivestronger banks in purchase and assumption opera-tions.89 Four large banks and three small banks re-ceived conditional approval for their rehabilitationplans and were requested to submit revised plans bythe end of July 1998.90 The three small banks willnot be allowed to engage in foreign exchange busi-ness. In any event, the resubmitted plans were ap-proved at the end of 1998: five of these banks havemerged (with public financial support) to create twonew banks, one has been recapitalized with fundsfrom the Bank of Korea and a foreign bank, and theremaining small bank has raised capital from currentowners (this capital was matched with public funds).

73

88The bridge bank was financed by the KDIC. In January, ittook over the deposits of the suspended merchant banks alongwith most of their performing assets. After a due diligenceprocess, the value of assets and liabilities transferred was set at8.7 trillion won and 12.1 trillion won, respectively. Shortly afterintervention, depositors were offered cash reimbursement, withhouseholds being compensated first, followed by enterprises andfinancial institutions. As of the end of June 1999, 99 percent ofprivate and institutional depositors had been repaid, as well as allfinancial institutions’call money deposits, for a total amount of9.7 trillion won. A further 4.4 trillion won remains to be repaid,mainly deposits of financial institutions.

89The five closed banks had capital adequacy requirements be-tween –4 percent and –11 percent at the end of March and a nega-tive net worth totaling 920 billion won (representing 7 percent oftotal assets of the banking sector).

90One small bank was given conditional approval despite re-ceiving a negative evaluation from the Evaluation Committee be-cause the Korean legislation did not allow for the closure of abank with positive net worth. This legislation was amended inAugust 1998.

APPENDIX II

All of these banks have signed memoranda of under-standing with the Financial Supervisory Commis-sion undertaking to meet recapitalization and re-structuring targets, such as branch closures.

The government has provided about 30 trillionwon (7 percent of GDP) to support commercialbanks. About 50 percent of this amount has beenused by KAMCO to purchase nonperforming loans,while the other half has facilitated mergers and thepurchase and assumption operations. Governmentrecapitalization of banks has been conditional on awritedown of current owners and managementchanges. As a result of the restructuring, the govern-ment now owns shares in 11 out of the 17 remainingbanks; its ownership exceeds 90 percent in 4 largebanks (Table 16).

Financial consolidation was helped by mergers andforeign investment. In September 1998 four rela-tively strong banks merged into two new entities,each undertaking facilitated by some KAMCO pur-chases of nonperforming loans. In the case of tworounds of mergers among much weaker partners, an-nounced in September 1998 and early 1999, govern-ment involvement in loan purchases and capital in-jection amounted to a total of more than 8 trillionwon and led to the effective nationalization of the twonew banks. Foreign investment of about $1 billionhas been announced, and several other significant un-dertakings are in advanced stages of preparation.

In the summer of 1998, banks not undercapital-ized at the end of 1997 were also subject to diagnos-tic reviews. As a result of these reviews and the Fi-nancial Supervisory Commission’s own appraisals,three small banks were subject to prompt correctiveaction procedures and were required to raise addi-tional capital.

Other financial institutions

The government has also recapitalized the special-ized and development banks, whose portfolios haddeteriorated significantly, and made them subject toregulations in line with those applied to commercialbanks. In the case of the Industrial Bank of Korea,these moves effectively reversed the process of pri-vatization that had begun before the crisis. Recapi-talization was done by buying equity in the banksand paying for it with government shares in publicenterprises. A total of more than 9 trillion won hasbeen injected into these banks.

Once the strategy for commercial and merchantbank restructuring was in place, the authorities tar-geted the restructuring of other financial institutionswhose soundness had deteriorated significantly be-cause of the severity of the crisis and inadequate su-pervision. At the end of March 1998, the FinancialSupervisory Commission estimated that this sectorhad about 30 trillion won in nonperforming loans,

74

Mergers of financial institutions and foreign invest-ment have been important elements in the changing fi-nancial structure in Korea. A number of mergers amongthe stronger banks have taken place, supported by pur-chases of impaired assets by KAMCO. These includethe merger between Kookmin Bank and Long-TermCredit Bank, and between Hana Bank and BoramBank, both announced in September 1998. Each mergerwas supported by about 300 billion won in nonper-forming loan sales to KAMCO.

Weaker banks were also encouraged to merge withgovernment support in the form of nonperforming loanpurchases and capital injection. These include theCommercial Bank of Korea and Hanil Bank, whichmerged in September 1998 to form Hanvit Bank; ChoHung Bank, Kangwon Bank, Hyundai Merchant Bank,and Chungbuk Bank, which announced their plans tomerge into one bank in early 1999. The government hasbecome the largest shareholder in both banks, althoughit intends to reduce its shareholdings by 2002, possiblyby selling stakes to strategic foreign investors. In bothmergers, memoranda of understanding were signed bymanagement, specifying targets on profitability, man-agement, and operational restructuring.

Foreign investment is also an important element inthe recapitalization of the Korean banking sector. InJune 1998, the International Finance Corporation in-vested $152 million in Hana Bank and $25 million inKLTCB. Germany’s Commerzbank invested $249 mil-lion in Korea Exchange Bank, acquiring a stake of 30percent, mainly by converting existing credits to thisbank into equity. In December 1998 a U.S. consortiumagreed, subject to due diligence, to purchase a 51 per-cent stake in Korea First Bank. In February 1999,HSBC Holdings similarly agreed to purchase a 70 per-cent stake in Seoul Bank. Conclusion of the deals ispending, with ongoing negotiations focusing on theevaluation and treatment of the nonperforming loans.In April 1999, Goldman Sachs announced that it wouldinvest $500 million to acquire a 17 percent stake inKookmin Bank. In the same month, Shinhan raised$400 million through global depository receipts. Mostrecently, ING Group acquired a 10 percent stake inHousing and Commercial Bank for about $280 million,while New York Life and the International FinanceCorporation signed memoranda of understanding topurchase a two-thirds stake in Kookmin Life Insurancefor $105 million.

Box 16. Korea: Mergers and Foreign Investment in the Financial Sector

Appendix II

about 7 percent of those institutions’total assets (andalmost equal to nonperforming loans in commercialbanks). In June 1998 the Financial SupervisoryCommission announced a restructuring plan for lifeinsurance companies, investment trust companies,leasing companies, and securities companies.

The restructuring program for life insurance com-panies began with diagnostic reviews of 18 “weak”life insurance companies by the Financial Supervi-sory Commission. As a result, four companies wereclosed through purchase and assumption-type opera-tions, seven were required to submit rehabilitationplans, while the remaining seven were required tosign memoranda of understanding with the FinancialSupervisory Commission undertaking to meet ex-plicit restructuring objectives. In 1999, six of thesecompanies (having failed to implement their plans)were put up for sale by the Financial SupervisoryCommission. To strengthen the supervision and reg-ulation of life insurance companies, the FinancialSupervisory Commission revised the solvency mar-gin regulation in April 1999, making it consistentwith the European Union standard. Accounting rulesand loan classification standards were also broughtin line with those applying to commercial banks.

In mid-1998, the Financial Supervisory Commis-sion revoked the licenses of two investment trustcompanies, and the six remaining were placed undermanagement improvement orders and required tosubmit rehabilitation plans and raise new capitalfrom owners and reduce their indirect borrowingsfrom trust assets.91 Measures to reform the sector in-

clude the requirement that all funds established aftermid-November 1998 be marked-to-market; further,all funds have to be marked-to-market by mid-2000.Ten out of 25 leasing companies were closed andtheir businesses transferred to a bridge leasing com-pany. Several securities companies were also closed,and regulations were issued to assure full segrega-tion of proprietary and customer accounts, whichhad been a problem.

The Financial Supervisory Commission also tooksteps to bring the regulation of mutual savings com-panies and credit unions into line with that of thecommercial banks. These companies were requiredin January 1999 to meet capital adequacy require-ments calculated on the same basis as commercialbanks. Companies that failed to meet minimumthresholds have been subject to prompt correctiveaction rules—resulting in the closure of 19 mutualsavings and 40 credit unions.

The Korean financial sector has thus been signifi-cantly consolidated (Figure 10). Since December1997, 9 out of 26 commercial banks and 19 out of 30merchant banks have been either closed or merged.Moreover, a substantial recapitalization effort hastaken place, including through foreign capital.

As of mid-1999, the government has spent closeto 47 trillion won (11 percent of GDP) to recapitalizefinancial institutions and purchase nonperformingloans (Table 17). The budgetary cost, that is, intereston bonds issued to financed restructuring, is close to2 percent of GDP. An additional amount of 16 tril-lion won (4 percent of GDP) in asset swaps has alsobeen spent to recapitalize government-owned, spe-cialized, and development banks.

Corporate restructuring

The process of reconstructing the financial sectorwas accompanied by measures to rehabilitate the fi-nances of many of the chaebols. A number of thelarger firms were dealt with by a modified version ofthe “LondonApproach.”92 This has involved setting

75

Table 16. Korea: Government Ownershipof Commercial Banks

Percent of Shares at theEnd of May 1999

Hanvit 94.8Cho Hung 91.8Seoul 93.8Korea First 93.8Korea Exchange 29.5Peace1 42.3Kookmin1 14.5Korea Housing & Commercial1 39.9Shinhan1 20.1KorAm1 34.7Hana1 46.2

Source: Financial Supervisory Commission (FSC).1Preferred shares.

92The “London Approach” is a framework for voluntary work-outs between creditors (banks) and borrowers (corporations). Theapproach involved establishing a corporate restructuring agree-ment, signed by financial institutions, under which they agreed tofollow specific procedures for debt workouts and to subject them-selves to binding arbitration by a private agency, especially set upfor the purpose, called the Corporate Restructuring CoordinateCommittee (CRCC). These procedures included the creation ofcreditor committees to deal with the restructuring of individualcorporations or conglomerates. Lead banks or groups of institu-tions holding more than 25 percent of a corporation’s debt wereable to call a creditors’committee meeting. An automatic stand-still on debt payments applied while the committee negotiated.Upon agreement among banks, the lead bank negotiated with thedebtor corporation. In all cases, arbitration by the CRCC has beenavailable to seek to resolve bottlenecks in the negotiations.

91One ITC subsequently converted into an investment trustmanagement company.

APPENDIX II

up debt workout units within banks and voluntarycreditor committees (including banks and othercreditors). These arrangements are expected to con-tribute to strengthening credit analysis and risk man-agement capacity within domestic banks. Debtworkouts often involved rescheduling, interest ratereduction, and debt forgiveness, as well as swaps ofbank debt for equity or convertible bonds. In thecase of small- and medium-sized enterprises, bankswere encouraged to roll over debt until the viabilityof such firms could be better assessed in the light ofthe performance of the economy. The largest fivechaebols are proving a less tractable problem, and itmay not be clear for some time to what extent fur-ther value adjustments to bank claims are necessaryin these cases. The entire process will not be con-cluded for some time and, in the meantime, the needfor further restructuring in the financial sector, in-cluding the provision of more government funds,may be needed.

Banks have provided some of the impetus for cor-porations to restructure. One reason for this is be-cause measures to strengthen prudential regulationsand supervision, including limits on large exposures

and connected lending, require banks to reduce theirexposures to corporations. Another reason is the roleof government-owned banks (commercial or devel-opment), which are better able to pressure largechaebols into restructuring.

Strengthening the Banking Environment

To prevent the recurrence of banking system prob-lems, the financial reform strategy calls for improv-ing the supervision and management of banks. Themain elements comprise a shift to consolidated banksupervision and a strengthening of prudential regula-tions and supervision; the liberalizing of restrictionson foreign ownership and management of banks;and a strengthening of the banks’credit evaluationand risk management capabilities.

As noted earlier, the Financial Supervisory Com-mission is responsible for the supervision and pru-dential regulation of all financial institutions. Regu-lations are being phased in to bring merchant banksand other financial institutions under the supervisory

76

26 commercialbanks1

(40%)

3 specialized anddevelopment banks (20%)

230 mutualsavings (10%)

8 investment trustcompanies (10%)

33 life insurancecompanies (9%)

12 commercial banks (19%)

5 state-ownedcommercial banks2

(18%)

210 mutualsavings (7%)

29 life insurancecompanies (7%)

1600+ credit unions (5%)

1600+ credit unions (5%)

11 merchant banks (4%)

30 merchant banks (6%)

4 investment trustcompanies (20%)

3 specialized anddevelopmentbanks (20%)

Financial sector at the end of 1996(share of assets)

Financial sector at the end of July 1999(share of assets as of March 1999)

Figure 10. Korea: Progress in Financial Sector Restructuring

Source: National authorities; and IMF staff estimates.1Does not include trust accounts.2Banks with majority government ownership (Hanvit, Cho Hung, Korea First, and Seoul).Also includes the Bank of Korea-owned Korea

Exchange Bank.Note: As of the end of July 1999, five commercial banks, 17 merchant banks, four life insurance companies, three investment trust companies, and

more than 50 mutual savings and credit unions had been closed.The state had intervened in 11 commercial banks. Four commercial banks and twomerchant banks had exited through merger. Categories are not mutually exclusive.

Appendix II

umbrella and to subject these institutions to pruden-tial standards in line with those applied to commer-cial banks. Supervision is also being enhanced tocover the full range of banking risks of financial in-stitutions on a consolidated basis.93

The authorities strengthened prudential standardsand supervision procedures—with special emphasison the regulations of foreign exchange activities—tobring them in line with best practice as set out in the“Core Principles for Effective Banking Supervision”recommended by the Basel Committee. On June 30,1998, the authorities introduced new loan classifica-tion standards and provisioning rules under whichloans more than three months overdue will be classi-fied as substandard, and the general provisioning re-quirement was increased from 1 percent to 2 percent.The Financial Supervisory Commission also intro-duced regulations to require the provisioning for se-curities losses and to cease the inclusion in Tier 2capital of all provisions for nonperforming loans (asof January 1999). The Financial Supervisory Com-mission is now preparing to issue guidelines for loanclassification and provisioning that would apply tobanks’end of 1999 accounts. The new guidelines aredesigned to take into account a borrower’s future ca-pacity to repay in classifying and provisioning loans.This prospective reform is likely to lead to a need forincreased provisions.

The Financial Supervisory Commission has an-nounced the strengthening of prudential supervisionand regulation of foreign exchange operations by

commercial and merchant banks.94 Regarding for-eign exchange liquidity management, compliancewith the guidelines that require short-term assets(less than three months) to cover at least 70 percentof short-term liabilities, and long-term borrowing(more than three years) to cover more than 50 per-cent of long-term assets, has been enforced for com-mercial banks as of January 1999 and will be en-forced for merchant banks as of December 1999. Amaturity ladder approach, monitored by the Finan-cial Supervisory Commission on a monthly basis,has been implemented for commercial banks sinceJanuary 1, 1999 and since July 1, 1999 for merchantbanks.95 Overseas branches and subsidiaries are in-cluded in the calculations. In addition, banks have tomaintain overall foreign currency exposure limitsper counterparty, including foreign currency loans,guarantees, security investments, and offshore fi-nance. In line with international best practice, thelimit on spot foreign exchange transactions of bankshas been removed, leaving a global limit on spot andforward positions.96

77

Table 17. Korea: Public Cost for Financial Sector Restructuring(as of mid-1999)

Percent of GDP Billions of U.S. Dollars

Recapitalization including outlays for deposit guarantee 6 22

Purchases of nonperforming loans or capital for asset management company 5 17

Interest cost (on budget) 2 7

Total 13 46

Memo item: Asset swaps 4 16

Source: National authorities; and IMF staff estimates.

93In particular, all trust accounts with guarantees are being re-garded as on-balance sheet items for supervisory and accountingpurposes. For capital adequacy ratio calculations, assets in suchaccounts are weighted at 50 percent as of January 1999 and willbe weighted at 100 percent as of January 2000. The operations offoreign affiliates are also now fully consolidated.

94The government has amended the Foreign Exchange Act tomove responsibility for the setting of foreign exchange open posi-tion limits and the supervision of foreign exchange risk to the Fi-nancial Supervisory Commission from the Bank of Korea, withinformation provided regularly through the Bank of Korea.

95This will require banks to report maturity mismatches for dif-ferent time brackets (sight to seven days, seven-day to one month,one to three months, three to six months, six months to one year,and over one year), and to maintain positive mismatches for thefirst period. From sight to one month, any negative mismatchshould not exceed 10 percent of total foreign currency assets, andfrom sight to three months, it should not exceed 20 percent ofsuch assets.

96This measure was designed to induce an increase in turnoverin the foreign exchange market, and enhance the pricing of theforward market so as to reflect interest rate differential. It has al-ready encouraged greater swap market activity, leading to acloser link between won and dollar money markets.

APPENDIX II

The large exposure of banks to the big conglomer-ates has been a major source of difficulty. Exposurelimits to single borrowers and groups were too gen-erous, and the authorities passed legislation in Janu-ary 1999 to tighten them. These limits have been re-defined to include all off-balance sheet exposures.Both single borrower and group limits for commer-cial and merchant banks will be progressively re-duced to 25 percent of total capital by 2004, and ag-gregate exposures in excess of 10 percent of totalcapital will be gradually reduced to 500 percent oftotal capital. Connected lending by merchant bankswill be limited to 25 percent of equity capital by Jan-uary 2001. Excesses in aggregate exposures andconnected lending will be published regularly.

Increasing foreign ownership and management ofbanks is recognized as a means to help recapitalizebanks, increase competition, and improve the man-agement of banks. Since December 1997, full for-eign ownership of merchant banks has been allowed.Nonresident purchases of equity in banks and otherfinancial institutions (excluding merchant banks) are

subject to the laws governing equity ownership ofKorean companies by foreigners. Currently, the indi-vidual foreign ownership limit is set at 50 percent,although the Financial Supervisory Commissionmay grant exemptions. (These limits are more gener-ous than those for residents, which are 4 percent fornationwide banks and 15 percent for regionalbanks.) In addition, the government has submittedlegislation to abolish regulations prohibiting for-eigners from becoming bank managers.

Banks have also been encouraged to adopt opera-tional improvements. Banks’rehabilitation planshave included specific benchmarks to improve theirprofitability and the quality of their portfolios.Banks have lowered labor costs and reduced theirstaffing, and have closed down uneconomicalbranches domestically and overseas.97

78

97The number of bank employees decreased by 34 percent andthe number of branches decreased by 17 percent as of the end of1998, compared to the end of 1997.

Despite Malaysia’s long history of outward-oriented macroeconomic and financial policies,

the country did not escape contagion from the Thaicrisis.98 In addressing the situation, Malaysia hasadopted some measures significantly different fromthose in the other countries, most notably the reintro-duction of capital controls in September 1998. WhileMalaysia has not adopted an IMF-supported pro-gram, it has nevertheless received technical assis-tance from bilateral and multilateral sources, includ-ing the IMF.

Background


At the outset of the crisis, Malaysia’s macroeco-nomic performance was good. Growth was about 8percent a year, driven in part by very high savingsrates. Prudent financial policies had kept inflationlow at about 3 percent, and there was a history of fis-cal surpluses. The current account deficit had peakedin 1995 at about 8!/2 percent of GDP, but had fallento slightly above 5 percent in 1996. External debt ap-peared manageable. The debt to export ratio wasabout 38 percent (roughly one-third of the Thairatio). Short-term debt comprised about 60 percentof international reserves, again lower than in theother crisis countries.

Beginning with the emergence of the financialcrisis in Thailand in mid-1997, Malaysia experi-enced increasing turbulence in financial markets (in-cluding a sharp increase in offshore ringgit interestrates). Market concerns about the economic vulnera-bilities in Malaysia were reflected in a sharp fall ofthe ringgit and the stock market in the order of 40percent to 50 percent by the end of 1997. Concernsincluded the current account deficit, rapid growth indomestic demand led mainly by large infrastructure

projects, strong credit growth, and the future courseof asset prices and their potential impact on thebanking system.

The authorities responded to market pressuresinitially through foreign exchange market interven-tion accompanied by an increase in interest rates,and by subsequently allowing the ringgit to depreci-ate. There were substantial outflows of capital,through commercial banks and sales of stocks byforeigners. By the first half of 1998, real output(year by year) had declined by 5 percent, led bysharp declines in manufacturing and constructionoutput. Sharp decreases in imports (by more than 25percent) exceeded the slowdown in exports (byabout 10 percent), leading to current account sur-pluses. Gross international reserves fell to 3.5months of imports at the end of 1997 from 4.6months in 1996. Severe corporate and financial sec-tor distress also led to a surge in the number oflisted companies seeking court protection fromcreditors. Nonperforming loans soared to more than25 percent by late 1998.


At the beginning of 1998, the Malaysian finan-cial system, which is dominated by the bankingsystem, comprised 22 domestic and 13 foreign-owned commercial banks (with about 69 percent ofthe system’s assets), 39 finance companies (about22 percent), 12 merchant banks (6 percent), sevendiscount houses (3 percent), and money and ex-change brokers. Nonbank financial institutions in-cluded the National Savings Bank, pension andprovident funds, insurance companies, and special-ized credit agencies (Table 18). The range of activ-ities that finance companies and merchant bankscould undertake had been gradually extended withsimilar, but not identical, regulatory require-ments—though there remains a sharp demarcationin the roles and activities of these institutions vis-à-vis commercial banks. A two-tier structure, pred-icated on market share and condition, was intro-duced at the end of 1994 for commercial banks andextended in 1996 to merchant banks and finance

Appendix III Malaysia

79

98This chapter draws on an initial draft by David Marston andcontributions from Patrick Downes, Michael Moore, and InciÖtker-Robe.

APPENDIX III

companies.99 There were restrictions on foreignownership (30 percent) and on the activities offoreign banks. In addition to the domestic financialsystem, there was an offshore market in Labuan,governed under its own separate legal framework bythe Labuan Offshore Financial Services Authority.

Precrisis Weaknesses in the Financial Sector

In the wake of market turbulence and contagioneffects in the second half of 1997, concerns amongmarket participants about the true condition and re-silience of the financial system increasingly becamea central issue, highlighted by the known frailtiesamong finance companies.


Standard official indicators of the Malaysian fi-nancial system soundness improved significantly in

the 1990s. The ratio of nonperforming loans to totalloans in banks and finance companies fell from 20percent in 1990 to about 3.8 percent for banks, and4.7 percent for finance companies in 1996; risk-weighted capital adequacy requirements rose to lev-els in excess of minimum Basel standards; and gen-eral provisions of commercial banks increased from0.75 percent in 1990 to 2 percent of outstandingloans in 1996.100 However, the persistent pace ofcredit expansion at an annual rate of nearly 30 per-cent (particularly by smaller Tier 2 financial institu-tions) to the private sector,101 in particular to theproperty sector and for the purchase of stocks andshares, exposed the financial system to potentialrisks from price declines in property and other assetsthat occurred in 1997.As a result, nonperformingloans rose in 1997, and official estimates showedfurther deterioration in the months leading up to theannouncement of the reform program. In response tothese concerns, Bank Negara Malaysia had earlier

80

Table 18. Malaysia: The Structure of the Malaysian Financial System(December 31, 1997)

Non-performing

Assets Deposits Loans Loans1_____________________ ____________________ ____________________ ___________Billions Percent Billions Percent Billions Percent Percent

Type of Institution of ringgit of total of ringgit of total of ringgit of total of loans

Financial institutions 699.0 100.0 433.4 100.0 402.0 100.0 n.a.Commercial banks (35) 481.4 68.9 300.5 69.3 276.4 68.8 4.5

Tier 1 (10) 314.2 44.9 199.0 45.9 168.5 41.9 n.a.Tier 2 (25) 167.2 23.9 101.5 23.4 107.8 26.8 n.a.

Finance companies (39) 152.4 21.8 106.5 24.7 102.6 25.5 7.8Tier 1 (3) 47.2 6.8 36.9 8.5 28.7 7.1 n.a.Tier 2 (36) 105.2 15.1 70.4 16.2 73.9 18.4 n.a.

Merchant banks (12) 44.3 6.3 26.4 6.1 23.1 5.8 3.4Tier 1 (4) 29.3 4.2 16.7 3.9 14.3 3.6 n.a.Tier 2 (8) 15.0 2.1 9.7 2.2 8.8 2.2 n.a.

Discount houses (7) 20.9 3.0 . . . . . . . . . . . . n.a.

Source: Bank Negara Malaysia.1Under the classification system in place during 1997.

99The two-tier regulatory system was perceived to have failed toachieve the desired strengthening of the capital base of domesticbanks and was abolished in April 1999. Until then, banks with atleast 1 billion ringgit in equity, and which complied with Bank Ne-gara Malaysia’s CAMEL requirements, were eligible to be Tier 1banks. These banks were given certain privileges (e.g., allowed tooperate foreign currency accounts on behalf of exporters, partici-pate in equity derivatives, undertake securities borrowing or lend-ing, and enjoy less stringent restrictions on branch expansion).

100Interpretation of the indicators of improved soundnessshould be treated with caution because, in an overheated econ-omy with an asset price bubble and very rapid credit growth andextensive collateralization of lending, the nonperforming loanratio may decline even though underlying asset quality deterio-rates. The deterioration only becomes obvious after the asset bub-ble has burst. Moreover, capital adequacy requirements are notadjusted for provisioning deficiencies.

101Rapid credit growth among smaller institutions in part re-flected their efforts to build their asset basis to achieve Tier 1status.

imposed limits on property lending (20 percent oftotal loans) and securities lending for the purchase ofshares, effective from April 1, 1997. In October1997, the authorities issued a directive that no newloans should be approved for the property sector, ex-cept for low income housing. The authorities alsoasked banks to submit credit plans for 1998 to mod-erate loan growth of the banking system to 20 per-cent by March 1998 and to 15 percent by the end of1998. Nevertheless, overall credit growth remainedstrong, falling only slightly, to 26.5 percent by theend of 1997.

In addition to the exceptionally high level of in-debtedness, the combination of the economic slow-down, decline in asset values (particularly propertyand stock market securities), rising interest rates,and the depreciation in the ringgit severely affectedcredit performance and bank profitability. Tight li-quidity conditions and segmentation of the interbankmoney market also contributed to narrowing interestspreads, especially for finance companies, and agrowing level of nonperforming loans in many fi-nancial institutions.


After the banking problems of 1985–87, the au-thorities took actions to improve the legal and regu-latory framework for banking supervision.102 In1989, a new banking law was adopted. In 1994, anew law on the Bank Negara Malaysia was enacted,and a large number of prudential regulations and cir-culars were issued covering enhanced prudential su-pervision, regulation standards, and the provision ofregular statistical reports and inspections.103 Thenew banking law and the Banking and Financial In-stitutions Act (BAFIA) provided broad regulatoryenforcement and intervention powers to the supervi-sory authorities. Bank Negara Malaysia updated reg-ulations, and overall a flexible framework was de-veloped that permitted the authorities, by and large,to address prudential concerns.

One area of uncertainty in the implementation ofthe prudential framework related to the fact that theBAFIA provided broad exemption powers to theministry of finance—albeit formally at the recom-mendation of Bank Negara Malaysia—with regardto individual prudential regulations, such as lending

to connected parties,104 ownership of shareholdingsin banks, and large exposure limits.105 There are,however, no reports of systematic or widespread useof this power.

Impact of the Regional Crisis andInitial Responses

As regional uncertainties unfolded, concernsabout the true condition of the financial system in-creased. These concerns stemmed from the excep-tionally fast rate of growth of credit in the bankingsystem, the private sector’s high leverage (163 per-cent), the concentration of bank loans in real estatedevelopment and in financing share purchases, andthe decline in asset qualities given the slowdown ineconomic activity. These worries were manifested indeposit flights to quality assets and institutions. Al -ready vulnerable to liquidity shocks, given low ex-cess reserves and high loan/deposit ratios, conditionsin the money market were aggravated by changedperceptions of counterparty credit risks that resultedin severe money market segmentation and pressureon the payment system. Bank Negara Malaysia re-sponded initially by acting to recycle liquiditythrough Bank Negara Malaysia deposit placements,supported by the announcement on January 20, 1998of a general guarantee of deposits.There was no for-mal deposit insurance scheme in Malaysia.106

The authorities, with a view to strengthening thefinancial sector, announced several prudential mea-sures, effective January 1, 1998, including a require-ment for banks to classify loans as nonperformingwhen they were three months overdue; an accelera-tion of the classification of “doubtful loans” from 12months overdue to six months, and “bad loans” from24 months to 12 months; and a rise in the requiredgeneral loan-loss reserves from 1 percent to at least1!/2 percent. The requirements for booking interest-in-suspense were also tightened, so that banks werenow required to reverse unpaid interest out of in-come and record it in the interest-in-suspense ac-count. Banks were also required to report on a quar-

81

Appendix III

102During this episode, nonperforming loans grew to 30 per-cent, and institutions were restructured with government and cen-tral bank support.

103Prior to 1998, banks in Malaysia did not routinely reportdata on a consolidated basis. When activities of subsidiaries weretaken into account, a number of banks announced a sharp deterio-ration in profitability and solvency.

104Lending to directors, officers, and employees was prohib-ited, as well as to firms in which these persons have an interestgreater than 5 percent—as a partner, manager, agent, or guaran-tor. Also, firms in which managers, directors, or employees heldshares were excluded from receiving credit from the institution.

105BAFIA prohibited institutions from holding shares in anycorporation. Exceptions were made for shares held as security oras a result of foreclosure or as repayment or credit.

106At its peak in January 1998, Bank Negara Malaysia place-ments amounted to 34.7 billion ringgit or 13 percent of GDP. Thisfigure declined by 15.8 billion ringgit in mid-February/Marchconcurrent with a reduction of reserve requirements from 13.5percent to 10 percent.

APPENDIX III

terly basis the ratio of nonperforming loans brokendown into substandard, doubtful, and loss; loans bysectors; general and specific provisions for bad anddoubtful loans; as well as their risk-weighted capitalratios.107

As of February 1998, only 10 commercial banksand four finance companies had Tier 1 status. Profitsof the system varied greatly; four banks and eight fi-nance companies made losses or zero profits whileeight banks and 23 finance companies had rates ofreturn on assets below 1 percent. In part, due to themonetary management instrument framework usedby Bank Negara Malaysia (differential reserve re-quirements across institutions, limited reserve aver-aging, and differential liquid asset requirement withno averaging) but more so related to the fact thatsince 1994, private sector loan/deposit ratios had ex-ceeded 90 percent, the banking system operated withvery low excess reserves, suggesting very limitedability to react to adverse liquidity shocks. (Loan/de-posit ratios were 96 percent for banks and 101 per-cent for finance companies.)

On March 25, 1998, the authorities announced apackage of measures aimed at strengthening the fi-nancial sector. The measures focused on a broad-based strengthening of the regulatory and supervi-sory framework requirements for increaseddisclosure; strengthening the finance company sec-tor through consolidation into a smaller number ofcore companies; and preemptive recapitalization ofbanks. The measures also included initiatives to im-prove the framework for bank liquidity manage-ment and monetary operations, and are detailed inBox 17.

Design of the Restructuring Program

Strengthening the Prudential Framework

The authorities vigorously implemented the an-nounced program. By April 1998 targeted improve-ments to the loan classification and provisioningstandards and the reduction in single borrower limitsfrom 30 to 25 percent of capital had been completedand incorporated in the ongoing supervisory andregulatory framework. Moreover, in addition to itsinitial broad diagnostic review of the banking sys-tem in March 1998 that used stress tests (projectionsof capital adequacy adjusted for varying scenarios of

nonperforming loans, specific provisions, collateralvalues, and other measures of other balance sheetrisks), Bank Negara Malaysia grouped the banksinto three categories: sound banks, those on a sec-ondary watch list, and those on a primary watch list.A program of intensified surveillance was institutedthat involved reaching understandings with eachvulnerable bank on a rehabilitation program. To aug-ment its surveillance system, Bank Negara Malaysiainitiated a program to develop an early warning sys-tem with the assistance of the World Bank. It alsotook the initiative to provide daily information to themarket on its assessment of liquidity based on adaily forecast of factors affecting reserve money.108

All disclosure requirements were observed by theend of May 1998, and a prudentially based frame-work for assessing bank liquidity risks was intro-duced effective August 2, 1998.109

Finance Company Merger Program

In addition to the announcement that finance com-panies would be subject to higher risk-weighted cap-ital adequacy requirements, the Malaysian authori-ties announced a program of consolidation of theindustry. The program envisaged the consolidationof 39 companies into eight, to be achieved throughthree modules. The first, covering 14 institutions,would involve the consolidation of finance compa-nies into their parent banks; the second, relating tothree firms, would involve consolidation through aswap of shares; and the third was a straight mergerof 15 companies into six anchor institutions. Compa-nies consolidated in this last module would carry aone-year government guarantee on the net assetvalue arrived at during the due diligence process.During 1998, however, as the economy and loanquality deteriorated sharply, and amid concernsabout the possible open-ended nature of the govern-ment guarantee, the finance company merger pro-

82

107Through March 1998, no provisioning was required for sub-standard loans, 50 percent for doubtful, and 100 percent forbad/loss. The amount on which provisions were to be made wasnot based on collateral; by regulation, real estate collateral mustbe marked-to-market semiannually while security collateral isrevalued daily.

108Based on this estimate, Bank Negara Malaysia also indi-cated its intention to either supply or withdraw liquidity and in-vited price or volume bids depending on the intended signal.

109It was envisaged that the framework would run parallel to theexisting liquid asset requirement for a period of six months duringwhich banks would be required to comply with the existing17 percent ratio. Officials indicated that this would be appliedflexibly, as there might be cases where banks—which were as-sessed to have adequately adapted to the new framework—inter-nal information systems, recomposition of liquid holdings accord-ing to types of asset holdings, and forecasting capabilities—wouldbe allowed to operate on the new system alone. Bank NegaraMalaysia reserved the right to extend compliance with the liquidasset requirement beyond the six-month parallel run in those caseswhere it was assessed that a bank was not adequately prepared forthe new framework. In early May 1999, 17 institutions migratedto the new framework.

gram was drastically scaled back. The sole remain-ing requirement was that finance companies affili -ated with banks be merged. Moreover, the initialschedule for compliance with the requirement for in-creased capital for the merged entities was post-poned, and the government guarantee of net assetvalues, aimed at facilitating purchase, was with-drawn. Smaller finance companies not eligible forsupport through DanamodalNasionalBerhad (therestructuring agency) would not receive governmentsupport for recapitalization. Notwithstanding theslower than anticipated progress in consolidatingthis industry segment, Malaysia reports that thenumber of finance companies declined by eight in1998 and is expected to decline by another 14 in1999.

Commercial Banks: Asset Management and Recapitalization

Complementing earlier initiatives to restore soundintermediation, Malaysia established in August 1998an institutional framework to strengthen efforts torehabilitate the commercial banking system by usingpublic funds to acquire nonperforming loans and re-capitalizing commercial banks. It also institutedmeasures that were focused on facilitating the re-structuring of corporate debt.

Institutional Framework

DanahartaNasionalBerhad, the public asset man-agement company, was established in June 1998 as apublic company under the Companies Act. TheDanaharta Act of 1998 gave Danaharta the ability toacquire nonperforming loans through statutory vest-ing and to appoint special administrators who cantake control and manage the assets of a borrower un-able to pay its debts. Danamodal was established inAugust 1998 as a limited liability company whollyowned by Bank Negara Malaysia, with objectives toinject new capital in undercapitalized banks and fa-cilitate rationalization of the system. The generalpolicy positions of both Danamodal and Danahartaincorporate principles of economizing the use ofpublic funds and finding least-cost solutions to gov-ernment. Although Danamodal issues progress re-ports on its operations, it has not released a publicstatement as to its operational principles, criteria forselecting operations, formulae for burden sharing, orexit strategy. Danaharta’s principles of operation, in-cluding valuation methodology, asset acquisitionguidelines, and expected nonperforming loan salesstrategy are already public knowledge. The Corpo-rate Debt Restructuring Committee has also been es-tablished to act as an informal debtor/creditor brokerto achieve debt restructuring as an alternative to

83

Appendix III

Loan classification and provisioning standards.Classification standards (including three months fornonperforming loans) to be brought to best practice; 20percent provisioning requirement against uncollateral-ized portions of substandard loans; off-balance sheetitems incorporated in the loan classification and provi-sioning system.

Capital adequacy framework. Increase risk-weightedcapital adequacy requirements of finance companiesfrom 8 percent to 10 percent, with interim requirementof 9 percent; minimum capital for finance companiesincreased from 5 million ringgit to 300 million ringgit;compliance with capital adequacy requirement requiredevery financial quarter.

Other prudential guidelines. Single borrower limitreduced from 30 percent to 25 percent of capital funds;prudentially based liquidity framework introduced;technical study on international practices regarding de-positor protection schemes to be undertaken.

Disclosure. Aggregate statistics on nonperformingloans, provisions, and capital positions for all financialinstitutions to be published monthly by the Bank Ne-gara Malaysia; all institutions to report and publishkey indicators of financial soundness on a quarterlybasis.

Intensified monitoring. More intensive and rigoroussupervision of banks through monthly stress tests byBank Negara Malaysia and a requirement for similarexercises by individual institutions on the basis of para-meters set by Bank Negara Malaysia.

Merger program.Bank Negara Malaysia-facilitatedmerger program of finance companies on market-basedcriteria. Mergers allowed only if merged entity wouldbe fully capitalized. Approval contingent on up-frontdue diligence. Government to extend a one-year guar-antee on net asset value determined from due diligence,and to share in upside gains as well as downside risks.Any institution electing to stay out of the mergerprocess was required to demonstrate its ability to com-ply with new capital requirements, failing which, ap-propriate action as provided by BAFIAwould apply.

Monetary measures. Commercial bank liquiditymanagement to be improved through widening the bandfor reserve averaging and providing market participantswith daily information on Bank Negara Malaysia oper-ations and liquidity forecasts; normal liquidity supportoperations to be separated from emergency liquiditysupport, and lending operations collateralized; the roleof base lending rate and maximum interest spread re-strictions, and liquid asset ratio, to be reviewed.

Box 17. Malaysia: Measures Announced on March 25, 1998

APPENDIX III

companies filing for bank bankruptcy under theCompanies Act. To facilitate coordination at both thepolicy and operational levels, an overarching Steer-ing Committee on Restructuring is chaired by theGovernor of Bank Negara Malaysia and composedof the managers of Danaharta, Danamodal, and theCorporate Debt Restructuring Committee.

Operational Process

In practice, the operations of Danaharta andDanamodal are guided by Bank Negara Malaysia’sclassification of banks. Bank Negara Malaysia ini-tially classifies institutions into a primary and sec-ondary watch list. The process is fine-tuned byDanamodal with the assistance of two investmentbanks acting as its advisors.110 Danamodal has identi-fied 14 institutions that are either currently undercap-italized or likely to be so in the future.111 Danahartaused the same database to identify 18 institutions (in-cluding the Danamodal 14) that are most likely tosell nonperforming loans. Danamodal’s intentionsare carefully coordinated with nonperforming loansales to Danaharta through ensuring thatDanamodal’s due diligence, and Danaharta’s nonper-forming loan acquisition and write-downs of capitalof affected institutions occur concurrently.

Danaharta has divided its nonperforming loan ac-quisition process into three stages, starting with se-cured loans, properties, and quoted shares. A secondstage will cover unsecured loans, to be followed byacquisitions of foreign currency loans. ThroughMarch 1999, Danaharta had acquired 23 billion ring-git in nonperforming loans from 37 financial institu-tions. The average discount at which Danaharta haspurchased nonperforming loans has been about 60percent of the principal value. As of the end of June1999, none of the acquired assets has been sold. OnJuly 1, Danaharta began the process of auctioning$143 million (face value) of foreign loan assets.

Danamodal has injected 6.2 billion ringgit as freshcapital into 11 financial institutions that representapproximately one-fifth of the financial system’sneeds. The Danamodal investment initially was asTier 2 subordinated debt that will be converted intoequity, debt, or a hybrid capital instrument.

The Corporate Debt Restructuring Committee hadbeen set up to prevent companies from abusing theprotection against creditors afforded by Section 176of the Companies Act. The process of resolution canbe initiated by either party, but once an appeal has

been filed, there is a six-month moratorium on ac-tion during which credit committees will work withthe affected parties to achieve a workout strategy. Asof the end of June 1999, 52 applications, represent-ing about 22 billion ringgit in affected debt, havebeen received by the Corporate Debt RestructuringCommittee. Creditor committees had been formed toconduct due diligence studies and to formulate re-structuring proposals, and these are in various stagesof completion. Four cases have been rejected andturned over to Danaharta for resolution.


As of the end of March 1999, Danaharta andDanamodal have spent about 15 billion ringgit(5 percent of GDP) to purchase nonperforming loansand recapitalize banks (Table 19).

Managing Credit Recovery and Rehabilitation

The reforms described above were complicated byconcurrent attempts by the authorities to stimulatecredit growth in the face of the economic downturn.By mid-1998, the authorities were faced with a sharpdecline in credit. Credit flows were negative to theextent of 1.5 billion ringgit, and undrawn commit-ments declined from 144 billion ringgit to 125 billionringgit between December 1997 and July 1998.While the Malaysian authorities wished to use inter-est rates to stimulate credits, they were constrainedby onshore/offshore interest rate differentials. Whilerelatively stable—and in fact declining in mid-1998—these differentials limited the scope for reduc-ing onshore interest rates without possibly triggeringringgit flight offshore. The authorities, therefore, in-troduced additional measures in September 1998aimed at eliminating the offshore ringgit market, fix-ing the exchange rate, and improving the conditionsfor increased bank lending. These measures included:

• Exchange control measures. A broad range ofmeasures to restrict international capital flows,which effectively eliminated the offshore ringgitmarket and prohibited nonresidents from repa-triating portfolio capital held in Malaysia for aperiod of 12 months, were adopted.112

84

110Danaharta employs the services of Arthur Andersen, whilethe financial advisors to Danamodal are Solomon Smith Barneyand Goldman Sachs.

111Two institutions have since been taken out of the Danamodalprocess.

112In view of continued weakness of investor confidence andconcerns about the possibility of a massive capital outflow uponthe expiration of the 12-month period in September 1999, the au-thorities in February 1999 replaced the 12-month rule with a de-clining scale of exit levies. Under this arrangement, for funds thatentered Malaysia before February 15, 1999, repatriation of theprincipal of portfolio investments is subjected to a levy, with thelevy decreasing with the duration of investment in Malaysia(starting from 30 percent, reduced in steps to 0 percent if repatri-ated after 12 months from the date of entry or September 1, 1998,whichever is later).

Appendix III

• Exchange rate policy. Subsequent to the intro-duction of capital controls, the authorities fixedthe exchange rate of the ringgit at 3.8 ringgit tothe U.S. dollar.

• Reserve requirements. Statutory reserve re-quirements were reduced to 4 percent of eligibleliabilities—from 10 percent in February 1998 to8 percent in August—applied to the mainte-nance period beginning September 15, 1998.The allowed averaging of 20 percent was keptunchanged.

• Loan limits. Existing limits on lending to theproperty sector, and for the purchase of shares,were relaxed. Loans for residential mortgagesup to a value of 250,000 ringgit were made ex-empt from the 20 percent ceiling on loans to theproperty sector—the previous exemption was150,000 ringgit. The ceiling on loans for sharepurchase was increased for commercial banksfrom 15 percent of portfolio to 20 percent. Also,financing margins for the purchase of motorvehicles were increased from 70 percent to85 percent.

• Credit floor target. A floor target for creditgrowth during 1998 was established at 8 per-cent. However, the authorities indicated thatweak banks on the Bank Negara Malaysia

watch list would not be expected to comply, andbanks that could indicate prudential con-straints—for example, rising nonperformingloans or the prospect of violating prudentialstandards—would be exempted from the target.

• Base lending rate. The formula for computingthe base lending rate was modified to reflectgreater sensitivity to Bank Negara Malaysiapolicy rates, and to lower the premium thatbanks charge.

• Loan classification. Bank Negara Malaysia an-nounced changes in the classification system sothat the default period for classifying loans asnonperforming was increased from three to sixmonths, and 20 percent specific provisions onsubstandard loans were no longer required.113

Effective March 24, 1999, Bank Negara Malaysiaamended its loan classification and provisioningguidelines again so that substandard loans—thoseoverdue between three and six months—would re-quire provisioning of 20 percent. Doubtful loans—overdue between six and nine months—would re-quire provisioning of 50 percent; and loss loans—those overdue for more than nine months—wouldrequire provisioning of 100 percent.114 Provisionsmust cover losses in the value of securities. Nonper-forming loans can be reclassified as performingwhen repayments are made continuously for sixmonths, rather than 12 months as required before.

85

Table 19. Malaysia: Public Cost forFinancial Sector Restructuring(As of the end of March 1999)

Percent Billions ofof GDP U.S. Dollars

Recapitalization including deposit guarantee 2 1.6

Purchases of non-performing loans 3 2.4

Total 5 4.0


113The modifications introduced in September 1998 should beseen against the more stringent loan classification system insti-tuted in December 1997. This system had introduced an auto-matic 20 percent provisioning against substandard loans definedas nonperforming for a period of three months. On the basis ofthat scheme, Bank Negara Malaysia classified institutions, andDanamodal and Danaharta determined estimates of recapitaliza-tion needs, and likely nonperforming loan sales. On that basis,nonperforming loans in July had increased to 14.2 percent.Banks, if they wish, can retain the more conservative threemonths’nonperforming loan classification. Of Malaysia’s 78 fi-nancial institutions, 21 (accounting for 46 percent of total loans inthe system) have retained this tight classification rule.

114For trade finance loans, the default periods for the classifica-tions are shortened as follows; substandard—default between oneand two months; doubtful—default between two and threemonths; and bad—default greater than three months.

The Philippines has been less affected thanneighboring countries by the Asian crisis and

did not undergo a similar degree of capital outflows,banking sector distress, and other forms of financialupheaval.115 Nevertheless, the combination of initialweaknesses in the banking sector, the fallout fromthe crisis in neighboring countries, and the possibil-ity of more intensive contagion led the Philippines toundertake preemptive measures to address potentialfinancial sector weaknesses. The program, devel-oped with the support of the IMF and the WorldBank, complements and deepens the efforts to im-prove the efficiency of financial intermediation de-ployed during the previous years.116

Background


The performance of the Philippine economy priorto the current crisis had been improving. After sev-eral years of accelerating growth rates, the expan-sion of real GNPreached 6.8 percent in 1996. Infla-tion remained well below 10 percent a year. Thebalance of payments remained strong, and foreignreserves of the Philippines’central bank, the BangkoSentral ng Pilipinas increased from less than $5 bil-lion at the end of 1993 to about $8 billion (1.7months of imports) at the end of 1997; strong capitalinflows—which reached $8 billion in 1996—offsetthe rapid widening of the current account deficit toalmost 5 percent of GDPin 1996.

In early 1997, the Philippines were faced with alarge decline in the stock market and pressure on theexchange rate. From mid-1997, although to a lesserdegree than in other Asian countries, the regional

crisis led to a further substantial disruption in eco-nomic activity. Following the devaluation of theThai baht in early July, the Philippines experienced alarge decline in capital inflows, a further fall in thestock market, and pressures on the peso. The BangkoSentral ng Pilipinas initially intervened in support ofthe peso; these efforts, however, proved unsustain-able when about $2 billion in Bangko Sentral ngPilipinas reserves were lost in a few days. The pesowas allowed to float and fell from about 26 pesos perU.S. dollar at the end of June 1997 to a low of about44 pesos in September 1998, before stabilizing atabout 36 pesos per U.S. dollar. Interest rates—mea-sured by the return on the 91-day treasury bill—in-creased gradually from 10.5 percent in June 1997 toabout 19 percent by January 1998 as the authoritiesattempted to keep the depreciation of the peso undercontrol; they subsequently fell back to about 13.5percent by early 1999. The authorities’strategy thusinitially focused on tight monetary and fiscal poli-cies. But, as stabilization took hold, the stanceshifted gradually toward supporting recovery.

Real GNPgrowth reached 5.8 percent in 1997—reflecting very strong growth in the first part of theyear—but fell to only 0.1 percent in 1998, while atthe same time inflation increased to over 10 percent,and accelerated further in early 1999.117 Despite thesignificant contraction in capital inflows—the capitalaccount surplus fell to $900 million in 1997 and $400million in 1998—foreign reserves recovered during1998 owing to the rapid adjustment of the current ac-count of the balance of payments, which recorded asurplus of over 1 percent of GDPin 1998.


Private banks in the Philippines comprise univer-sal or “expanded” banks (by far the largestcomponent of the banking system), nonexpandedcommercial banks, thrift banks, and rural banks

Appendix IV The Philippines

86

115This appendix draws on an initial draft by Enrique de laPiedra and contributions from Elizabeth Milne and GretaMitchell-Casselle.

116Banking sector reform is a major component of the IMF’sStand-By Arrangement approved in March 1998, and the WorldBank’s Banking Sector Reform Loan, approved in November1998.

117In addition to contagion from the other Asian countries, thePhilippines was also adversely affected by the weather pattern as-sociated with El Niño.

Appendix IV

(Table 20).118 Only one universal bank, the Philip-pine National Bank, is partly owned by the govern-ment.119 Thrift banks cater mainly to the consumerretail market and small- and medium-sized enter-prises. There are two fully government-owned spe-cialized banks, the Land Bank of the Philippines andthe Development Bank of the Philippines, whichalso undertakes some commercial banking func-tions. The small government-owned Islamic bank iscurrently under a rehabilitation plan. Total assets ofthe banking system amounted to over 2.8 trillionpesos in 1998, roughly equivalent to annual GNP;banks represented 90 percent of the banking system,up from 85 percent in 1991.

There are significant restrictions on the size andoperation of foreign banks.120The number of foreignbanks that operate wholly owned branches in thePhilippines is currently capped at 14. Other foreignbanks may apply for a universal or a nonexpandedbank license and can operate by acquiring up to 60percent of the voting stock of an existing domestic

bank or of a new institution incorporated locally. Onaverage, foreign equity is about one-fifth of total eq-uity for universal banks, 13 percent for nonexpandedcommercial banks, and negligible in the case ofthrift and rural banks.121

Precrisis Weaknesses in theFinancial Sector

Prior to the Asian crisis, the Philippines had un-dergone a period of financial liberalization, coupledwith rapid financial deepening.122 The changes ap-pear to have come too fast for supervision to remainfully effective. Hence, although the financial condi-tion of the Philippine banking system in terms ofcapital adequacy was better than several of theneighboring countries, the corporate sector wasmore resilient, and the real estate boom less pro-nounced, there were notable weaknesses in the fi-nancial sector around the time of the crisis.123 The

87

Table 20. Philippines: Selected Banking Sector Indicators as of December 31, 1998(In billions of pesos, unless otherwise indicated)

Commercial Banks Thrift Rural Banking_____________________________________________Universal Nonexpanded2 Foreign Total Banks Banks1 System

Physical CompositionNumber of banks 21 20 12 53 117 826 996Total number of branches 3,519 650 11 4,180 1,357 1,114 6,651

Balance Sheet

Total Assets 2,029 217 310 2,556 210 58 2,824of which: Loan portfolio (net) 1,209 131 141 1,481 124 40 1,646

Deposit liabilities 1,318 140 128 1,586 132 38 1,755of which: Foreign currency 480 50 71 601 8 0 609

Capital 259 44 8 311 28 9 348

Annual Income and ExpensesTotal operating income 184 21 25 230 17 3 249

of which: interest income 166 18 20 204 15 2 221Net income before tax 22 0 4 27 0 0 27

Trust Assets 364 22 27 414 5 — 419

Source: Bangko Sentral ng Pilipinas.1Data for rural banks in the Balance Sheet and Annual Income and Expenses entries are for June 30, 1998.

118Universal and nonexpanded commercial banks are jointlyreferred to as the commercial banks.

119In June 1996, the government had reduced its ownership inPhilippine National Bank from 100 percent to 45.6 percent. Thestock in private hands is widely dispersed.

120Foreign banks are not authorized to open more than sixbranches or to borrow from head offices more than $4 for every$1 of domestically held capital.

121Legislation pending before Congress as of mid-1999 wouldallow for full foreign ownership of distressed banks.

122Earlier banking sector measures included significant govern-ment assistance to resolve a banking crisis in the early and mid-1980s and a recapitalization of the central bank in 1993.

123In contrast to other Asian countries, property demand wasflat during 1989–94 and in early 1997; the office vacancy ratewas only 2 percent. In early 1997, the Bangko Sentral ng Pilip-inas took steps to limit bank lending for real estate purposes.

principal weaknesses and supervisory challenges arediscussed next.


Several characteristics of the Philippine bankingsector that existed before the crisis appear to havemade it vulnerable to shocks. They include:

• A very rapid growth in banks’lending followinggradual liberalization of banking and substan-tial financial deepening.124 The speed of thisgrowth in lending brought about significantproblems, including a decline in the averageloan quality, an increase in unhedged foreignexchange operations, and for some banks, grow-ing involvement in real estate lending. Theseproblems have been particularly acute forsmaller commercial banks, thrift banks, andrural banks.

• A significant growth in financial intermediationin foreign currency. Mirroring overall trends inbanking sector activity, financial intermediationin foreign currency also grew significantly untilmid-1997, owing in part to the prevailing signif-icant institutional advantages for bank opera-tions conducted in foreign currency.125 Totalforeign currency deposits in the banking systemexpanded at an annual rate of 38 percent from1994 to 1996, and reached a maximum of $16.3billion in June 1997; as a result, foreign cur-rency deposits increased to more than half oftotal bank liabilities, up from only 3 percent in1990.

• A weakening in capital levels.The deteriorationin banks’portfolios led to a significant weaken-ing in capital levels. Average capital levels re-mained still comfortable at the onset of the cri-sis, but capital adequacy ratios had been on adownward trend since 1993. Capital adequacydeclined from 19.2 percent in 1993 to 16 per-cent in 1997. Moreover, average capital levelsdisguised problems of individual institutions—one large and several among the smaller ones—that suffered from low capital levels relative tothe risk implied by their asset portfolios. Fi-nally, a special regime granted to a few univer-sal banks allowed them to operate with a mini-mum capital adequacy requirement of 8 percentinstead of the 10 percent generally required.

Weaknesses in Prudential Regulation andSupervision

Despite significant progress since the early 1990s,at the onset of the crisis the banking sector still suf-fered from problems both in the supervisory frame-work and the implementation of supervision. Themain areas of concern were shortcomings relative tointernational norms of prudential standards, the ef-fectiveness of bank supervision, and the enforce-ment of the bank regulatory and supervisory frame-work. Finally, mechanisms for bank exit wereinadequate.

Prudential standards

Problems in bank supervision included thefollowing:

• Loan classification and provisioning.Until mid-1997, loan-loss provisions were not required forbank assets backed by collateral, even if suchassets were classified as substandard, or for anybank assets classified as “especially men-tioned.”126 There was also limited capacity ofthe Bangko Sentral ng Pilipinas to assess bankrisks.

• Marking to market. Philippine accounting con-ventions regarding banks’portfolio investmentsfollow closely the U.S. pattern. Banks were notrequired to mark to market their equity portfo-lios, but in contrast to U.S. norms, this includedtheir trading accounts—thus they could deferrecognition of losses. The large changes in thevalue of banks’securities portfolios followingthe onset of the crisis resulted in significant

88

APPENDIX IV

124Total deposits and loans of the banking system increased by27 percent and 35 percent a year, respectively, during 1993–96,and expanded a further 11 percent and 16 percent, respectively,during the first six months of 1997. Financial intermediationdeepened significantly in the Philippines, as the ratio of moneyand quasi-money to GDPincreased from 34 percent in 1991 to 61percent in 1997, while the ratio of banking system claims on theprivate sector to GDPgrew from 18 percent to 56 percent duringthe same period. Nevertheless, the degree of financial deepeningin the Philippines remained among the lowest of other Asiancountries affected by the regional crisis.

125Following a long period in which interest from foreign cur-rency operations was fully tax-exempt, it is now subject only to a7.5 percent withholding tax compared to 20 percent in the case ofpeso deposits. At the same time, profits from the banks’foreigncurrency deposit unit operations are taxed at a 10 percent prefer-ential rate on gross income, compared to the standard tax rate of35 percent on net profit from other operations. Moreover, domes-tic banking activity, including resident foreign currency depositunits, is subject to the gross receipts and documentary stamptaxes, while transactions in foreign currency with nonresidentsand with other foreign currency deposit units are exempt. Finally,while peso deposits are subject to reserve requirements, mostlyunremunerated, foreign currency deposits are not subject to re-serve requirements.

126The “especially mentioned” classification was routinelyused for loans where supervisors found minor documentaryweaknesses, including for third party guarantors.

Appendix IV

changes in their market value, but the adjust-ment to book values arising from marking tomarket was not properly accounted for. The truesoundness of certain banks was thus not readilyapparent.

• Transparency. Although banks listed in thestock exchange were required to disclose theirbalance sheets publicly on a quarterly basis, noinformation was provided on the level of non-performing loans, classified assets, and loan-loss provisions, for instance. Unlisted bankswere under no obligation to disclose their finan-cial accounts to the public. External audits ofbanks were not integrated into the supervisoryprocess, as external auditors were not under anobligation to report materially adverse factorsand events to banking supervisors.

• Consolidated supervision.The Bangko Sentralng Pilipinas did not address bank solvency on aconsolidated basis, nor did it consolidate thebanks’trust activities with on-balance sheet ac-tivities to determine capital adequacy. It also didnot attempt to consolidate foreign currency ex-posures between banks and their affiliated for-eign currency deposit units in determining com-pliance with prevailing regulatory limits.

Effectiveness of supervision

Although some improvements took place duringthe years prior to the beginning of the Asian crisis,supervisory practices in general remained weak.

• On-site examinations.The on-site examinationprocess was not oriented toward an analytic ap-proach that would allow an assessment of risk atthe bank level and of the systems used by banksto manage risk.127 Rather, it was heavilyweighted toward the examination of asset qual-ity and compliance with laws, rules, and regula-tions. Moreover, bank solvency was assessed onan unconsolidated basis. Finally, the schedulingof on-site bank examinations was too con-strained. On-site examinations were undertakenon an annual basis, and additional examinationsrequired Monetary Board approval. Bank se-crecy legislation prevented supervisors from ac-cessing disaggregated loan information, pre-venting assessments of risk concentration.

• Off-site monitoring. The approaches used foroffsite monitoring were not well-suited to eval-uate the financial condition of the institutions.The CAMELsystem excluded sensitivity analy-sis for market risk. The individual component

ratings were applied mechanistically with norole for the supervisor’s judgment; in addition,to determining the overall CAMELrating of abank, the supervisors normally averaged the in-dividual component ratings; as a result, it waspossible for an insolvent bank to be evaluated as“good.”

Enforcement capacity

Even though the supervisory authorities generallyreceived good compliance from banks on routine re-quests, their ability to require banks to adopt actionsto correct problems was limited. Furthermore, penal-ties for noncompliance with norms and regulationswere low, and there was inadequate protection forsupervisors against lawsuits related to actions takenin the course of their official duties. There were alsono mandatory administrative actions to be triggeredwhen a bank’s financial condition deteriorated be-yond certain preestablished limits.

Framework for bank exit

The framework for bank closures was time-consuming. If the Monetary Board deemed that atroubled bank could not be rehabilitated, it would beclosed and placed under Philippine Deposit Insur-ance Corporation (PDIC) receivership. However,there were no specific time limits for the MonetaryBoard to take such a decision. The PDIC had to de-termine within a maximum of 90 days whether to re-habilitate or close the bank; it was limited in its roleas a receiver since it could not dispose of assets of abank in receivership until the bank was closed. Evenworse, the Secrecy of Bank Deposits law preventedthe PDIC from gaining access to information regard-ing the liabilities covered by deposit insurance be-fore it was named as receiver of the bank. All localcurrency bank deposits are insured up to 100,000pesos. The actual liquidation of a bank may take sev-eral years to complete under the current system, re-flecting inadequacies in the judicial system. More-over, even when fully under way, in some cases thecourts have ordered the authorities to reverse theirdecision and reopen a bank. Since supervisors, aswell as the management and board of the BangkoSentral ng Pilipinas, do not have explicit immunityfrom prosecution (although such immunity has beenproposed in legislation pending before Congress), anoverruling by the judiciary has had a chilling effecton the Bangko Sentral ng Pilipinas’s willingness toact.

The process of closing down a bank has beencomplicated further by the fact that the Bangko Sen-tral ng Pilipinas may incur financial losses. Suchlosses may arise from uncollateralized overdraft

89

127In early 1998, a pilot test of risk-based supervision was initi-ated in one bank.

lending to a troubled bank; the inability by theBangko Sentral ng Pilipinas to realize the collateralbacking emergency loans to banks (in part owing tovaluation problems); and the fact that, in the event itruns out of resources, the PDIC can borrow from theBangko Sentral ng Pilipinas.

Avoiding a Crisis and Bolstering theFinancial Sector

The Philippine authorities adopted a financial sec-tor reform program in early 1998 aimed at strength-ening the banks’capacity to withstand shocks, en-hancing the prudential and supervisory framework,and encouraging market-based consolidation in thebanking sector. Minimum capital and loan-loss pro-visioning requirements were tightened, regulatorystandards in a number of areas were brought closerin line with international best practice standards, thefocus of bank supervision was reoriented toward theanalysis of risk, and bank exit policies werestrengthened. The main provisions of the financialsector reform program include the following.128

Bank Capital Requirements

Minimum capital requirements for banks wouldbe raised through the end of 2000, with intermediateminimum levels for the end of 1998 and the end of1999.129 At the same time, the lower capital ade-quacy requirement of 8 percent (instead of the usual10 percent) that was allowed for certain universalbanks was phased out in January 1999. The largerbanks have not encountered problems in meeting thenew minimum capital requirements. Several entitiesamong the other groups of financial institutions haveencountered more problems meeting the new re-quirements. The authorities have addressed the moresevere cases by stimulating mergers, signing memo-randa of understanding with the Bangko Sentral ng

Pilipinas, and setting a clear timetable for compli-ance or applying the policies contained in the matri-ces of sanctions and prompt corrective action ap-proved recently by the Monetary Board.

Loan-Loss Provisions

Banks are now required to make a general loan-loss provision of 2 percent and specific loan-lossprovisions of 5 percent for loans especially men-tioned and 25 percent for secured substandard loans.The new specific provisions had to be met by April1999 and the general provision by October 1999. Toavoid an undue restriction of credit growth, theBangko Sentral ng Pilipinas in early April 1999 tem-porarily freed any net increase of bank lendingabove the March 1999 level from the general provi-sioning requirement. Most institutions are comply-ing with these new regulations. For those that arenot, especially thrift and rural banks, the sanctionsoutlined in the newly instituted policy of prompt andgraduated corrective actions, such as restrictions onbranching and payment of dividends, are applied.

Marking to Market

The Bangko Sentral ng Pilipinas has requiredbanks to start marking to market their trading securi-ties portfolio. Accounting standards have beenagreed upon with the Bankers Association of thePhilippines, and the relevant price benchmarks havebeen discussed with market operators.

Transparency and Disclosure

To enhance transparency and market discipline,the Bangko Sentral ng Pilipinas instructed all bankslisted in the Philippine Stock Exchange to disclosepublicly, as of December 1998, detailed informationon a quarterly basis, including the level of nonper-forming loans and the ratio of nonperforming loansto the total loan portfolio, the amount of classifiedassets and other risk assets, and the extent of specificand general loan-loss reserves.

Consolidated Supervision

The authorities have started to supervise financialconglomerates in a consolidated fashion. Since leg-islative changes are needed to fully impose consoli-dated capital requirements and extend consolidatedsupervision of financial institutions to include theirinterests in nonfinancial ventures, the authoritieshave proposed an amendment of the General Bank-ing Act in this regard. The amendment is currentlyunder consideration by Congress. The Bangko Sen-tral ng Pilipinas has already begun to consolidate

90

APPENDIX IV

128The Philippine National Bank, the only large bank in finan-cial difficulties, is not fully subject to the regulations set outbelow, but the authorities have reconfirmed their intention to pri-vatize the Philippine National Bank by mid-2000. In the mean-time, the realization of losses has brought down the capital ade-quacy requirement to about half of the required 10 percent. Anexternal audit is in progress that will be the basis for privatizationand increase in capital adequacy requirement.

129Minimum capital is to increase to 5,400 billion pesos (from3,000 billion pesos) for expanded commercial banks and to 2,800billion pesos (from 1,625 billion pesos) for commercial banks.Rural and thrift banks will continue to have different require-ments depending on the region they will operate in. The maxi-mum capital requirement for thrift banks will be 400 millionpesos (from 200 million pesos) and for rural banks 32 millionpesos (from 20 million pesos).

Appendix IV

limits on foreign currency exposures between banksand their affiliated foreign currency deposit units.

Bank Licensing

Stricter licensing guidelines for establishing bankshave been in place since July 1998, focusing on thethree additional requirements: the submission of thestatement of income and expenses for the last threeyears for each of the subscribers; evidence of assetownership; and in the case of a foreign bank, certifi-cation by the home supervisory authority that itagrees with the proposed investment.

Supervisory Methods

The Bangko Sentral ng Pilipinas has changed thefocus of its supervision activities from a purely com-pliance-based and checklist-driven assessment ofbanks’condition to a forward-looking and risk-basedframework. Together with the change in emphasistowards a risk-based approach, significant improve-ments have taken place in the area of rating method-ologies. The CAMELrating system has been revisedto ensure that the composite rating will never be bet-ter than the bank’s individual factor rating for capitaladequacy. As of July 1998, “sensitivity to marketrisk” (“S”) was added to arrive at a CAMELS ratingsystem, and the composite rating system will bebased on the weighted sum of the component rat-ings, with each component assigned a differentweight depending on the size, complexity of activi-ties, and risk profile of the institution being rated.

The Role of External Auditors

External auditors of banks have been required,since the end of September 1998, to report to theBangko Sentral ng Pilipinas all matters that couldadversely affect the financial condition of theirclients, any serious irregularity that may jeopardizethe interests of depositors and creditors, and anylosses incurred that substantially reduce the bank’scapital. Noncompliance by auditors with this re-quirement will lead to loss of accreditation.

Bank Resolution

The authorities also addressed problems in therecognition and resolution of weak banks:

• Intensified bank monitoring. The Bangko Sen-tral ng Pilipinas has adopted a program of inten-sified monitoring of selected banks and is con-ducting special examinations of banks withoutspecific previous authorization on the basis of aregularly updated list of banks in potential dis-tress, which it started to compile based on for-

ward- and backward-looking indicators in early1998.

• Bank receivership. The authorities have agreedto adopt two measures to improve the ability ofPDIC to act as the receiver of banks, includingselling assets of distressed banks to pay for theadministration costs related to receivership, andfaster approval by the Monetary Board of a pro-posed liquidation.

• Prompt corrective action for bank capital short-falls. The Bangko Sentral ng Pilipinas hasadopted explicit procedures to be used whenbanks fail to reach certain thresholds of capitaladequacy. In particular, the authorities have is-sued a matrix of sanctions and of graduated cor-rective actions to be taken according to the de-gree of capital shortfall and noncompliance withother prudential norms.

• Measures to reduce Bangko Sentral ng Pilipinasfinancial losses. The Philippine authorities havetaken several steps to reduce the financial riskfor the Bangko Sentral ng Pilipinas associatedwith assistance to banks in distress. A samplingincludes stopping the distribution of dividends,imposing certain obligations on the banks’own-ers and officers, and eliminating uncollateral-ized overdrafts.

Bank Restructuring

Measures in the area of bank restructuring are cur-rently limited to the privatization of the remaininggovernment stake in the Philippine National Bankand the announcement of merger incentives:

• An important element of the authorities’strat-egy to enhance the overall soundness of thebanking system is to strengthen the financialperformance of Philippine National Bank, thesecond largest bank in the country in terms ofassets. Philippine National Bank’s financial per-formance has deteriorated significantly due tothe effects of the current regional crisis. In early1999, the bank retroactively recognized the im-pairment of its assets and wrote down its capitalaccordingly. Philippine National Bank’s risk-weighted capital fell below regulatory norms,although it continued to meet the minimum cap-ital requirement. The authorities intend to priva-tize their remaining holdings of the bank stockby mid-2000.

• The Monetary Board has granted additional in-centives for banks considering the option ofmerging if otherwise unable to comply with thenew minimum capital and provisioning require-ments. The merged bank would in general be al-lowed time to comply with specific rules andregulations.

91

Linkages with Corporate Sector Reforms

Philippine companies have been affected by theregional crisis to a lesser degree than those in neigh-boring countries, in part because of their lower expo-sure to foreign debt and better corporate perfor-mance. However, there is a concentrated ownershipstructure in the corporate sector as well as the crossshareholdings between banks and corporations. Theauthorities’structural reform agenda includes somemeasures to strengthen the corporate sector, includ-ing a strengthening of the institutional capacity ofthe Securities and Exchange Commission to dealwith distressed corporations while protecting thecontractual rights of creditors. The new emphasis onrisk-based bank supervision methods as well as themore rigorous disclosure requirements for bankswill also provide an early warning system of corpo-rate distress.

Impact of the Regional Crisis on Bank Performance

In spite of early measures to avoid a wider bank-ing crisis, the banking sector underwent a period ofincreased stress. The developments include thefollowing:

• Activity in the banking sector decelerated. Inthe first half of 1998, total deposits expandedmerely by 5 percent, and the stock of outstand-ing loans contracted by over 2 percent. The fi-nancial crisis seemed to have prompted a moreconservative attitude on the part of the bankswhile noticeably slowing the demand for credit.At the same time, banks started to scale downtheir expansion programs to ensure compliancewith new minimum capital and loan-loss provi-sioning requirements that became effective inthe fourth quarter of 1998.

• Foreign currency deposits contracted, falling by17 percent between June 1997 and June 1998.Most of the Philippine banking sector’s foreigncurrency liability exposure is to domestic resi-dents, which may make it less vulnerable tocapital flight than in other Asian countries. Non-residents’deposits grew by 70 percent a yearduring 1994–97 and continued growing in thefirst half of 1998, but still accounted for lessthan 25 percent of total foreign currency de-posits by June 1998.

• The decline in the banking sector’s capital ade-quacy requirements prior to the crisis has beenreversed; it reached 17.6 percent at the end of

1998.130 On average, all classes of banks en-joyed healthy capital adequacy ratios; amonglarge banks, all except one have a capital ade-quacy ratio well in excess of 10 percent. Theimprovement in the level of capital adequacy ra-tios reflected a slowdown in banks’assetgrowth, a shift in asset composition from loansto investment in government paper, and otherzero risk-weighted assets, as well as the newminimum capital requirements.

• Nevertheless, asset quality deteriorated. Therapid increase in the ratio of nonperformingloans to total loans since mid-1997 is an impor-tant indicator of the growing problems in thebanking sector. The officially reported overallnonperforming loan ratio reached 14 percent bymid-1999, compared to 4 percent in June 1997.The nonperforming loan ratio is much higheramong thrift banks and rural banks than it isamong commercial banks.

• Bank earnings fell, reflecting a general slow-down in the business environment and the needto constitute the new loan-loss provisioning re-quirements. The envisaged tax deductibility ofloan-loss provisions was not implemented asscheduled, hence the after-tax profit of bankswas much weaker than expected, making itmore difficult for some smaller banks to complywith minimum capital requirements. For thebanking system as a whole, the average returnon equity declined to 1 percent by the end of1998, from 1.7 percent in 1997, and an averageof over 2.3 percent from 1990 to 1996.

• Bank failures.Since the start of the current diffi -culties, one small commercial bank, seven thriftbanks, and 18 rural banks have failed. Theircombined assets amounted to less than 1 percentof total assets.

Unlike the crisis countries, however, the structureof the financial system has seen little change: onesmall commercial bank was closed and several oth-ers are in talks to merge.131 Reflecting the lesserproblems in the financial system, no public fundshave been needed for financial sector restructuring,except for liquidity support provided to the onesmall commercial bank that was subsequentlyclosed.

92

APPENDIX IV

130This may not fully reflect the impact of the new minimumcapital and loan-loss provisioning requirements, which only be-came mandatory as of the last quarter of 1998.

131This excludes closures of thrift and rural banks, which areroutine.

The floating of the Thai baht on July 2, 1997 trig-gered a deep macroeconomic and financial sec-

tor crisis.132 Thailand was not only the first amongthe crisis countries, but the first to enter into a stabi-lization program supported by the IMF.

Background


Under the framework of a pegged exchange rateregime, Thailand had enjoyed a decade of robustgrowth performance, but by late 1996 pressures onthe baht emerged. Pressure increased through thefirst half of 1997 amidst an unsustainable current ac-count deficit, a significant appreciation of the realeffective exchange rate, rising short-term foreigndebt, a deteriorating fiscal balance, and increasinglyvisible financial sector weaknesses.

Following mounting exchange rate pressures andineffective interventions to alleviate these pressures,the baht was floated on July 2, 1997.133 In light ofweak supportive policies, the baht depreciated by 20percent against the U.S. dollar in July. To arrest thepressure, Thailand, on August 20, 1997, entered intoa three-year Stand-By Arrangement with the IMF,augmented with funds from the World Bank, theAsian Development Bank, Japan, and other coun-tries. Key policy measures of the program includedsteps to restructure the financial sector, such as fiscaladjustment and the continuation of a floating ex-change rate system.

As the crisis spread to other Asian countries andthe sentiment of investors toward the region remainedshaky, rollover ratios of short-term debt declined andthe baht continued to depreciate. Economic activityalso declined, as investment, consumption, and ex-port demand fell sharply. The increasing lack of con-fidence in the government’s ability to manage the sit-

uation culminated in a change of government in No-vember 1997. After a significant strengthening of theeconomic program there was, starting in early 1998, agradual return of confidence, which was reflected in afirming exchange rate, although real output continuedto decline.

To support the exchange rate, monetary policywas initially tight, resulting in relatively high moneymarket rates. In March 1998, this was combinedwith a more accommodative fiscal stance to allowautomatic stabilizers to work. By mid-1998, with adeeper recession than expected and a markedstrengthening of the baht, monetary policy waseased and interest rates were gradually lowered. Byearly 1999, the exchange rate had stabilized andmoney market interest rates stood below precrisislevels. Growth for 1999 is now projected to be posi-tive, albeit small. Notwithstanding considerableprogress in financial sector restructuring, and an im-proved market sentiment, significant downside risksto economic recovery remain.


As of December 1996, the financial system inThailand consisted of 15 domestic commercialbanks, 14 branches of foreign banks, 19 foreignbanks established under Bangkok InternationalBanking Facilities, 91 finance and securities compa-nies, 7 specialized state-owned banks, some 4,000savings and agricultural cooperatives, 15 insurancecompanies, approximately 880 private providentfunds, and 8 mutual fund management companies.Total assets of the system amounted to 8.9 trillionbaht (190 percent of GDP), of which the commercialbanks alone accounted for 64 percent (121 percent ofGDP), finance companies for 20 percent (39 percentof GDP), and specialized state banks for 10 percent(8 percent of GDP). One large commercial bank andtwo finance companies were majority state-owned.

In an effort to make Bangkok an international fi-nance center that could compete with Hong Kong andSingapore, the Bangkok International Banking Facili-ties was established in 1993 to formalize offshorebanking business in Thailand. Bangkok International

Appendix V Thailand

93

132This chapter draws on an initial draft by Mats Josefsson andcontributions from Carl-Johan Lindgren.

133Thailand had committed almost all its foreign reserves in itsefforts to support the fixed-exchange rate regime.

Banking Facilities’operations mainly involved bor-rowing in foreign currencies from abroad and on-lending the funds locally and conducting internationaltrade financing. As an incentive for development, theBangkok International Banking Facilities benefitedfrom a number of tax exemptions. As of December1996, 45 financial institutions held licenses to operateoffshore banking businesses, of which 15 were Thaicommercial banks, 11 foreign bank branches alreadyoperating in Thailand, and 19 other foreign banks.

Precrisis Weaknesses in the Financial System


In response to banking sector weaknesses in the1980s, the Thai authorities had initiated reform mea-sures, including the creation of the Financial Institu-tions Development Fund, a separate legal entitywithin the Bank of Thailand with a mandate to re-structure, develop, and provide financial support(liquidity and solvency) to financial institutions.Notwithstanding these and other reform efforts, theauthorities failed to properly manage the risks in therapidly growing banking system. Structural weak-nesses that, in conjunction with the macroeconomicdevelopments outlined above, led to the emergenceof a full-blown banking crisis included:

• The quality of loan portfolios in banks and fi-nance companies was weak. In banks, nonper-forming loans (more than six months’overdue)were 7.2 percent of total loans at the end of1995, and increased to 11.6 percent in May1997. In finance companies, nonperformingloans also increased sharply over the first fewmonths of 1997 (from 6 percent at the end of1996 to 12 percent in May 1997). Given theweak accounting standards, market analysts be-lieved the figures were too low, and estimatednonperforming loans to be at least 15 percent oftotal loans for banks, and at least twice as highin finance companies. Most finance companies’assets were related to real estate.

• Banks and finance companies had not put asidesufficient reserves for their rapidly deteriorat-ing loan portfolios. There was, for example, noprovisioning requirement for substandard loans:loan classification and loss provisioning wasonly tightened in March 1997.134


The structural weaknesses resulted, in part, fromweaknesses both in the content and the implementa-tion of prudential regulations. The main weaknesseswere:

• The rules for loan classification, provisioning,and accounting were inadequate and were ap-plied inconsistently. Thus, the reported capitaladequacy ratios were grossly misleading sinceloans were not appropriately classified and pro-visioned for. For example, financial institutionshad built up large loan portfolios of increasinglyquestionable quality, secured by generally over-valued asset collateral. These loans were oftensimply restructured (“evergreened”) when pay-ment problems arose and not reclassified.

• Interest on nonperforming loans continued toaccrue and, hence, significantly overstated fi-nancial sector earnings. This had made it possi-ble to pay dividends, bonuses, and taxes onnonexistent profits, effectively decapitalizingthese institutions.

• There were no prudential limits on loan concen-tration. In the absence of such restrictions,banks built up excessive exposure to particularsectors such as the property market. There wasalso excessive lending based on collateral ratherthan proper credit assessment; thus, when theasset price bubble burst, banks faced rapid de-clines in the value of their collateral.

• The prudential framework was generally weakand fragmented. The ministry of finance wascharged with the overall authority for supervi-sion of banks and finance companies, but haddelegated the day-to-day responsibility for su-pervision to the Bank of Thailand. The ministryof finance had authority to grant, suspend, andrevoke banking licenses and to intervene inbanks and finance companies through a controlcommittee for each institution. Liquidation offinancial institutions was under ministry of fi-nance authority and was subject to the bank-ruptcy act for nonfinancial corporations.

Impact of the Crisis and InitialResolution Measures

The Crisis in Finance Companies

Finance companies had disproportionately thelargest exposure to the property sector and were thefirst institutions affected by the economic downturn.Much of the later spillover to other institutions, inparticular to the banking sector, was triggered by thefailure of the initial measures to stabilize this sector.

94

APPENDIX V

134In March 1997, the Bank of Thailand introduced provision-ing requirements against substandard loans in banks and financecompanies at 15 percent and 20 percent, respectively, to be imple-mented over a two-year period. However, under pressure fromweak institutions the authorities soon extended the implementa-tion period to five years.

Appendix V

Financial sector restructuring was initiated onMarch 3, 1997, when the Bank of Thailand and theministry of finance announced that 10 as yet un-named finance companies had asset quality prob-lems and insufficient liquidity, and would need to in-crease capital to cover weak real estate loans andfinance growth. According to the announcement,there was to be a tight deadline for the increase incapital—the shortest period of time allowed in thelaw—and, if the finance companies were unable toraise capital, they would have to sell their shares tothe Financial Institutions Development Fund. TheFinancial Institutions Development Fund wouldtherefore effectively take over the finance company.The public was also assured that, apart from the 10institutions mentioned, all other financial institutionscould increase capital through their own efforts.

In the period from March to late June, the Bank ofThailand—in absolute secrecy—provided liquiditysupport at below market interest rates to 66 financecompanies. There was also liquidity support to twobanks.135 A small part of the liquidity was providedfrom the Financial Institutions Development Fund’sown accumulated reserves, some through FinancialInstitutions Development Fund borrowing in theovernight repo market, and the remainder was fi-nanced through Financial Institutions DevelopmentFund bonds, which were purchased by the Bank ofThailand.

To stop the liquidity drain, on June 29, 1997, theBank of Thailand suspended for 30 days the opera-tions of 16 finance companies, including 7 of the 10“initially targeted” institutions, based on their capitalinadequacy and the need for liquidity. These 16companies were required to submit rehabilitationplans to the Bank of Thailand by July 11. Companiesthat failed to submit plans, or whose plans were re-jected by the Bank of Thailand/ministry of finance,would have their licenses revoked and be absorbedby Krung Thai Thanakit, a majority government-owned finance company. In addition, four otheridentified “core” finance companies were invited toplay a role in absorbing some of the suspended com-panies. The government also announced that it:

• would not suspend any more finance companiesbeyond the 16 already suspended; and

• would guarantee all domestic and foreign de-positors and creditors, of all finance companiesother than those suspended.

However, these measures failed to calm the mar-kets, mainly because there was increasing uncer-tainty over the exact extent of the guarantees due toinconsistencies in official statements. With inconsis-tencies among the different announcements unre-

solved, official assurances raised rather than reducedthe market’s apprehension.

Design of Initial Resolution Efforts

After the peg of the baht was abandoned on July2, 1997, the rapidly depreciating exchange rate andfalling property values led to major deterioration ofbanks’loan portfolios. This raised concerns aboutthe solvency of the entire financial system. Esti-mates undertaken by an IMF advisory mission inmid-July 1997, based on rough assumptions of thelevel of problem loans and recovery rates, suggestedthat by mid-1997, some 500 billion baht ($17 billionor 9 percent of projected 1997 GDP) would be re-quired to restore the legally required minimum capi-tal adequacy of banks and finance companies. The“hole” was estimated to be about 270 billion baht($9 billion) in banks and 230 billion baht ($8 billion)in finance companies. These aggregate figures indi-cated that many, if not most, of Thailand’s 91 fi-nance companies were insolvent, while banks as agroup were solvent, but undercapitalized. In the ab-sence of detailed supervisory information on indi-vidual institutions, there was an expectation that thefive largest banks (representing about two-thirds ofthe banking system) were among the strongest in thesystem, so that several of the small- and medium-seized banks most likely were insolvent.

At this stage, the authorities developed a strategyto restructure the financial system in cooperationwith the IMF. The strategy was based on three steps:

• exit of all nonviable financial institutions;• issuance of a temporary blanket guarantee pro-

tecting all depositors and creditors in the re-maining financial institutions, so as to calm themarket and give the authorities sufficient timefor restructuring measures; and

• restructuring and rehabilitating of the Thai finan-cial system to raise to international standards.

The strategy depended on the implementation of arange of supporting measures, in particular:

• macroeconomic policies that restored fiscal andmonetary control;

• establishment of the FRAto replace the Bank ofThailand and the ministry of finance temporar-ily as decisionmaker on all matters related to fi-nancial sector restructuring; and

• announcement of a consistent and comprehen-sive medium-term restructuring strategy, to beexplained to the public through a professionallymanaged information campaign.

Implementation

On August 5, 1997, the ministry of finance andthe Bank of Thailand issued a joint statement an-

95

135For some finance companies, the amounts involved were inexcess of four times capital.

nouncing measures to be taken to strengthen confi-dence in the financial system. The implementationof the program to a large extent followed the stepsjust outlined. The policy framework was built onthe separation of finance companies facing actual orimminent insolvency from those that were judged tobe viable, based on the following three criteria:(1) magnitude of liquidity support (borrowing fromFinancial Institutions Development Fund); (2) dete-rioration of capital; and (3) the size of nonperform-ing loans.

Finance companies deemed to be facing actual orimminent insolvency were suspended and given 60days to complete due diligence and present a reha-bilitation plan to the Committee on Supervision ofMerger or Transfer.136 In all, 58 finance companiesout of 91 had their operation suspended, includingthe 16 finance companies suspended earlier. At thisstage, the authorities decided not to take any actionagainst commercial banks, even though there weresigns that at least two small banks faced seriousweaknesses.

A government guarantee for depositors and credi-tors in all remaining finance companies and bankswas announced, and the conditions were spelled outin a joint press release by the ministry of finance andthe Bank of Thailand.137 However, there were majoruncertainties surrounding the guarantee, includingwhether it would be legal for the Bank of Thailandto provide the liquidity to honor the guarantee.Shareholders and holders of subordinated debt werenot covered by the guarantee.

There were delays in initiating some of the an-nounced restructuring measures. The magnitude ofthe problems, along with interagency issues of coor-dination and political problems, all took time to re-solve, delaying the implementation of necessarypolicies. The committee managing the process was

reorganized, and only by late September were in-structions issued to the suspended companies onwhat information they had to submit, and which cri-teria had to be met before they would be allowed toresume operations.

The government through its actions made clearthat it stood fully behind the Bank of Thailand. Tosupport the restructuring/rehabilitation process,emergency decrees were passed on October 24, 1997to: (1) establish the FRAto deal with the suspendedfinance companies; (2) amend the CommercialBanking Act and the Finance Company Act so as toempower the Bank of Thailand to request capital re-ductions, capital increases, or changes in manage-ment in troubled commercial banks and financecompanies; (3) establish an asset management com-pany to deal with assets of the 58 finance companiesthat had their operations suspended, or impaired as-sets in any financial institution in which the Finan-cial Institutions Development Fund had acquiredshares (intervened) and assumed management con-trol; and (4) amend the Bank of Thailand Act to em-power the Financial Institutions Development Fundto lend to these institutions with or without collat-eral, raise the fee charged to financial institutionswhose depositors and creditors were protected, andmake explicit the government’s financial support ofthe Bank of Thailand (thus making the guaranteemore credible).

The finance companies were given until the end ofOctober 1997 to submit their rehabilitation plans tothe FRA, which in turn had one month to assess therehabilitation plans submitted and to make a recom-mendation to the ministry of finance on how manyfinance companies should be allowed to resumetheir operations. With the assistance of private con-sultants, the FRArefined its analysis to allow thor-ough technical consideration of the rehabilitationplans based on uniform criteria.

On December 8, 1997 the FRAand the ministry offinance announced the permanent closure of 56 fi-nance companies. Only two companies were al-lowed to reopen, subject to the stipulated new capitalbeing paid in. The fact that various politically attrac-tive but financially dubious merger proposals hadbeen rejected was a sign of the integrity of the analy-sis.138 Immediately after the decision to close the 56companies, the focus of the FRAshifted to manag-ing their assets, in order to prevent a further deterio-ration in asset quality prior to disposition.

96

APPENDIX V

136The committee had initially been set up to handle the sus-pension of the 16 finance companies.

137The treatment of depositors and creditors in the 16 initiallysuspended and 42 later suspended finance companies was guidedby the following principles: (1) depositors in the 16 finance com-panies had a choice between receiving payment under terms andconditions specified by the company in question, or to takepromissory notes issued by Krung Thai Thanakit, with conditionsand interest varying according to the size of deposits; (2) credi-tors in the 16 finance companies were not protected—they couldshare in the proceeds of the liquidation process, in line with thepriority of their claim; (3) depositors in the 42 finance companieswere given essentially the same option as depositors in the other16 finance companies—with the exception that the promissorynotes were to be issued by Krung Thai Bank and were negotiable;(4) creditors in the suspended 42 finance companies were allowedto swap their claims for notes issued by the Krung Thai Bank,under the same condition as depositors, although the interest ratewas set at 2 percent.

138The closure of the 56 finance companies required the coordi-nated action of some 5,000 officials and security personnel.

Appendix V

Banking Sector Issues

Magnitude of the problem

Initial crisis resolution had focused solely on fi-nance companies, but it soon became evident thatbanks also faced problems and that the public hadstarted to lose confidence in the banks. Banks’assetquality was deteriorating rapidly, and many banks ex-perienced deposit withdrawals and reduced rolloverrates for external credit lines; these were facilitatedthrough liquidity support from the Financial Institu-tions Development Fund/Bank of Thailand. Seven outof the 15 commercial banks were facing such severeliquidity problems that they needed liquidity supportfrom the Financial Institutions Development Fundmore or less on a daily basis. The deposit withdrawalsrepresented a flight from private banks perceived tobe weak into larger stronger domestic banks, branchesof foreign banks, and state-owned specialized banks.

Based on Bank of Thailand data as of the end ofJune 1997, assessments of the financial condition ofThai banks were made. These assessments indicatedthat all banks had a capital shortfall (that is, a capitaladequacy ratio below 8.5 percent); the shortfallamounted to roughly 400 billion baht if internationalbest practices were to be applied, and approximately250 billion baht on the less stringent Bank of Thai-land criteria.

Resolution principles

On October 14, the authorities announced theirstrategy for dealing with the banking sector, basedon the following principles:

• All banks were required to adjust (first writedown and then increase) capital in order to ab-sorb the losses that had already occurred. Theyhad to meet the new and more stringent rules onloan classification and provisioning that wouldshortly be issued by the Bank of Thailand;

• None of the banks was allowed to pay dividendsfor the remaining part of 1997 or during 1998;

• The Bank of Thailand was to initiate discus-sions with each individual bank on how thebank would be recapitalized. Following suchdiscussion, banks were required to present re-capitalization plans to the Bank of Thailand;

• The requested capital would have to be injectedat the latest during the first quarter of 1998;

• If a bank could not raise the capital requestedwithin the time given, the Bank of Thailandwould have the right to demand a memorandumof understanding with the bank that would giveexisting owners additional time to provide thecapital, as long as they could furnish the Bankof Thailand with viable, legally binding, plansfor recapitalization;

• Banks were encouraged to try to find foreignpartners since it was unlikely they could raiseall the capital needed in Thailand;

• For banks that could not raise the capital, losseswould be written off against capital, ensuringthat existing shareholders lose their stake, withthe exception of a token position that had to bekept for legal reasons. The Financial InstitutionsDevelopment Fund would then take control ofthe banks, recapitalize them and later privatizethem (by selling them to domestic or foreign in-vestors) or merge them with another bank; and

• The strategy clearly stated that no bank wouldbe closed and that depositors and creditorswould be fully protected by the governmentguarantee.

Banking sector intervention

The Bank of Thailand was initially reluctant tointervene in banks since it feared that interventionscould cause a run on the whole banking system andseverely deepen the crisis. With the October 1997amendments to the Commercial Banking Act, theBank of Thailand was given specific powers towrite down capital and change management in trou-bled commercial banks. Based on these new powerson December 31, 1997, the Bank of Thailand inter-vened in a medium-sized bank, Bangkok Metropol-itan Bank: management was changed, and theBangkok Metropolitan Bank was told that if capitalhad not been raised within roughly three weeks, theBank of Thailand would order the bank to reduce itscapital and have the Financial Institutions Develop-ment Fund recapitalize it. In the following month,the Bank of Thailand intervened in two more small-or medium-sized banks—First Bangkok City Bankand Siam City Bank. These three banks represented10 percent of banking system deposits. In mid-May,the Bank of Thailand intervened in seven financecompanies; the government thus had become theowner of six banks and nine finance companies,which accounted for roughly one-third of totaldeposits.139

The authorities hired a financial advisor to assistthem with developing a strategy to deal with thethree banks mentioned above and with BangkokBank of Commerce. This financial advisor workedout an initial proposal for the banks to be absorbed

97

139In addition to the banks mentioned above, government-owned banks include Bangkok Bank of Commerce (a privatebank that had earlier failed and been taken over by the FinancialInstitutions Development Fund) and Radanasin Bank, a bank andan associated finance company that was set up by the governmentin early 1998 to buy good assets from closed finance companies.

by either domestic or foreign banks. Meanwhile, inthe course of 1998, these banks’assets continuouslydeteriorated to the point where nonperforming loanswere approaching 70–85 percent in each bank. Al -though clearly increasingly insolvent, no decisionwas made on how to proceed, since more problemswere emerging and there was an urgent need for theauthorities to work out a comprehensive plan onhow to address other outstanding weaknesses in thefinancial sector.

Toward a Comprehensive and ViableFinancial Sector

The above measures primarily addressed the im-mediate liquidity and solvency problems. There wasalso a need for a strategy to strengthen the remain-ing viable financial institutions. Consequently an ef-fort was made to focus on the longer run and morestrategic needs for supporting a sustainable finan-cial sector. Measures included revamping the pru-dential framework, and taking steps to restructurethe sector.

Improved Prudential Framework

Following failed attempts to tighten prudentialrules in early 1997, the severity of the crisis helpedfocus attention on a fundamental revamping of theprudential framework. The strategy was to maintainthe capital adequacy requirement of 8.5 percent forbanks and 8.0 percent for finance companies, butallow gradualism in building up loan-loss provi-sions according to a fully transparent timetable.There was to be an immediate tightening of loanclassification and interest suspension rules, whileprovisioning requirements were to be increasedevery six months to bring them fully into line withinternational best practices by the end of 2000. Thisphasing was thought necessary to give the bankingindustry the required time to adjust and raise thenew capital.

New loan classification, loss provisioning, and in-terest suspension rules were issued on March 31,1998. Loans had to be classified into five categories,and strict rules on interest accrual were established.The new loan-loss provisioning requirements weregradually tightened by 20 percent every six monthsstarting July 1, 1998, thus making it possible for therequirements to be fully implemented by the end of2000. A new regulation was also issued on the valua-tion of collateral for loans above a certain size; itmust now be independently appraised. Finally, to setclear incentives for banks and finance companies toactively initiate restructuring of nonperformingloans,rules for how restructured loans should be

classified and provisioned for were defined in regu-lations for debt restructuring.140

A Comprehensive Plan for Financial SectorRestructuring

By mid-1998, the depth of the economic and fi-nancial sector crises raised doubts whether the mea-sures described above, combined with the strength-ening of prudential rules, would be sufficient tocreate the necessary private sector recapitalizationand efficient restructuring of the sector. In particular,the deteriorating domestic and regional economicenvironment, and the associated decline in assetquality, had introduced uncertainties about the effec-tiveness of the market-based and private-sector-ledrestructuring of Thailand’s financial system. Whileseveral private banks, including the two largest, hadbeen able to secure injections of private capital inearly 1998, and a controlling share in a small bankhad been acquired by a foreign strategic investor,their success had not been widely shared, and it wasdoubtful whether the banking system would be in acondition to support the government’s efforts to se-cure economic recovery.141

Hence, on August 14, 1998 the government an-nounced a comprehensive financial sector restruc-turing package to address the remaining weaknessesin the financial sector. The augmented restructuringstrategy focused on the following elements:

• creation of a high-level financial restructuringadvisory committee (FRAC) to advise the min-ister of finance and the governor of the Bank ofThailand;

• the commitment of public funds to assist in therecapitalization of viable banks and financecompanies;

• incentives for accelerating corporate debt re-structuring at an equal rate with recapitalization;

• the efficient management of nonperformingassets;

• the exit, merger, or sale of nonviable commer-cial banks and finance companies;

• equitable loss-sharing arrangements and con-tainment of public sector costs;

• the strengthening of prudential supervision andan accelerated adoption of international bestpractices;

• operational restructuring of state banks and theirpreparation for eventual privatization; and

98

APPENDIX V

140These regulations were based on rules used by the U.S. Of-fice of the Comptroller of the Currency.

141Most new subscribers were foreign institutional investors,and foreign ownership of the two largest private banks ap-proached 50 percent; foreign investors also acquired controllingshares in some finance companies.

Appendix V

• Bank of Thailand intervention in remainingnonviable institutions.

Commitment of Funds Through CapitalSupport Facilities

The objective of the package was to restore andmaintain the solvency and credibility of the Thai fi-nancial system. The government recognized that theterms and conditions of any public support facilityshould be transparent and subject to the strongestpossible safeguards, while facilitating the restructur-ing and consolidation of financial institutions andprotecting the economy from further weakening.Two schemes were designed to assist with the recap-italization of the viable financial institutions. Thefirst aimed at catalyzing new private Tier 1 capital.The second provided financial resources and incen-tives to accelerate corporate debt restructuring andencourage new lending through the provision of Tier2 capital (Box 18).

To facilitate and oversee banks’efforts to raisecapital, at the beginning of every six-month period,the Bank of Thailand makes an assessment of eachbank’s provisioning needs, profitability, and growthin assets. It then calculates whether the bank has suf-ficient capital to meet the required capital adequacyrequirement. If a bank falls short of capital, it wouldbe required to present plans on how, what amount,and by when capital would be raised, and those com-mitments would be formally agreed upon in a mem-orandum of understanding. Banks that cannot raisecapital through their own efforts would be recom-mended to apply for public support under the Tier 1scheme. The Bank of Thailand has reserved the rightto intervene in any financial institution that fails tomeet commitments under their memoranda of under-standing.

Facilitating the Establishment of Private AssetManagement Companies

Additional arrangements were needed to providemaximum flexibility for financial institutions todeal with their bad assets, while allowing them torestructure their balance sheets. Thus the govern-ment devised a plan under which institutions wereencouraged to set up, capitalize and fund privateasset management companies. Such companieswere defined as financial institutions, which allowsthem to borrow, relend funds to existing customers,and provide full flexibility in setting interest rates.However, asset management companies are not per-mitted to take deposits. The rules also provide for atransfer of assets from the founding institution to itsasset management companies without legal impedi-ments or taxes and fees. According to the rules,

asset management companies must be fully consoli-dated into the founding institutions’balance sheets,provided the institution owns more than 50 percentof the shares. Two large private banks have an-nounced that they will set up their own private assetmanagement companies, and others are consideringdoing so.

On August 14, 1998, the Bank of Thailand inter-vened in all remaining banks and finance companiesconsidered nonviable. This included two morebanks, one small- and one medium-sized, and fivefinance companies. Subsequently, the authorities de-cided to merge three of the intervened banks into theexisting state-owned banks (Krung Thai Bank andRadanasin Banks), and create a new bank (BankThai), from the merger of Krung Thai Thanakit, oneintervened bank, and the 12 intervened finance com-panies.

Five banks are now in the process of being priva-tized and are expected to be privatized by the end of1999. In July 1999, one additional small bank wasintervened. The government is supporting the priva-tization process through capital injections and by of-fering stop-loss guarantees and/or profit and losssharing arrangements to potential new investors.Four of the intervened banks have been closed: oneformally through liquidation, and the other three in-directly through mergers with state-owned banks.

Asset Disposal

The FRAhas now almost completed the disposalof assets from the 56 closed finance companies. Inits first auction, it managed to sell 36 billion baht ofassets (hire purchase loans, mortgages, and com-mercial loans) and in a second auction it sold 28 bil-lion baht of noncore assets (foreclosed assets, re-possessed vehicles, equity, and debt instruments).The recovery rate in these two auctions was 47–48percent. In a third auction in December 1998, FRAoffered 370 billion baht ($10 billion) of lower qual-ity business loans, mainly consisting of corporateloans with or without questionable collateral. Thisauction was less than fully subscribed and obtainedfar lower prices because of the low quality of theloans. In total, 32 billion baht of assets were sold,with a recovery rate of 37 percent, and an additional110 billion baht of assets were later allocated to bid-ders after profit-sharing arrangements had been ne-gotiated guaranteeing a minimum recovery rate ofabout 20 percent of book value. For the first time,the government asset management company estab-lished in late 1997 was allowed to participate in theauction as a bidder of last resort, and it ended up be-coming the owner of assets with a book value of185 billion baht for which it paid about 17 percent.In sum, out of the 860 billion baht in finance com-

99

pany assets managed by the FRA, 206 billion bahthave been sold to the private sector and 185 billionbaht to the asset management company. Total recov-ery to FRAis about 96 billion baht or 25 percent offace value; final recovery value will only be knownafter the outcome of profit sharing arrangementsand asset management company liquidations arecompleted.

Bank Privatization and Resolution

The private sector-led recapitalization of Thaibanks has worked well, and all seven of the remain-ing private banks, have raised sufficient capital forthe next 12 months. However, all banks need to raiseadditional capital before they can have sufficientcapital to meet the end-2000 loan classification andprovisioning rules.

Of the six state-owned banks, four are in theprocess of being privatized with the assistance offoreign investment banks. It is expected that the pri-

vatization process will be completed this year. Thegovernment has made significant efforts to ensurethat the process is transparent, and that there areproper profit/loss sharing arrangements in place toensure a deal can be justified politically. Thetimetable and strategy for the privatization of theother two state-owned banks is being reconsideredto ensure that the eventual sale of shares in thesebanks can substantially contribute to reducing thegovernment’s final restructuring costs.

The financial sector in Thailand has been consoli-dated although only one commercial bank has beenclosed (Figure 11).The number of domestic com-mercial banks has declined by two as a result ofmergers. The most significant change is with the fi-nance companies—56 were closed and a further 13(together with five banks) were merged. Some for-eign capital has been invested in commercial banksand finance companies, most notably two private do-mestic banks that now have foreign ownershipgreater than 50 percent.

100

APPENDIX V

These capital support facilities apply to banks and fi-nance companies incorporated in Thailand and deemedviable by the Bank of Thailand.

Tier 1 capital. Under this facility, the governmentmatches capital injected by private investors (includ-ing existing shareholders). Any institution that entersthe scheme is required to make full provisions upfront,in line with Bank of Thailand’s end-2000 loan classifi-cation and loan-loss provisioning (LCP) rules. Exist-ing shareholders must thus bear all the up-front losses,and the government in making its contribution willrely on due diligence performed by private investors.The Financial Sector Advisory Committee (FRAC) ar-bitrates on any disputes regarding the magnitude of thewrite-down of losses. To qualify for support, the insti-tution must present an operational restructuring planincluding measures to strengthen internal control andrisk management, increase revenues, cut costs, andstrengthen internal procedures for dealing with non-performing loans. These plans must be acceptable tothe FRAC and the Bank of Thailand. The governmentwill provide the capital injection in the form of equitypaid up with 10-year, tradable government bonds car-rying a market-related interest rate. The new govern-ment/private capital injections would have preferredstatus over existing shareholders. The government andthe new investors have the right to change the board ofdirectors and management of the institution. The gov-ernment also reserves the right to appoint board mem-ber(s) according to the size of its equity holding, andhas the right of nominating at least one board member

even if this is not warranted by the size of its equityshare.

Tier 2 capital. This scheme is set up to encouragecorporate restructuring. At the end of each quarter, theinstitutions can apply to the FRAC for government in-jection of Tier 2 capital by reporting any debt-restruc-turing agreement, the original loan contract, and evi-dence that the borrower had been able to service theloan. The amount of Tier 2 capital provided is set at aminimum equal to the total write-down exceeding pre-vious provisioning or 20 percent of the net increase inlending to the private sector. Each institution is eligibleto receive a maximum amount of Tier 2 capital injec-tion equal to 2 percent of risk-weighted assets. Withinthis limit, the amount of Tier 2 capital injection for newlending to the private sector cannot exceed 1 percent ofrisk-weighted assets. Also, no single debt restructuringagreement is eligible for more than 10 percent of theamount available to the institution. Institutions thatbring forward the end-2000 LCPrules will be allowedto defer the cost of debt restructuring over a period offive years (20 percent per year) while those whichchoose not to bring forward the end-2000 rules willhave to take the costs according to the existing regula-tion (full loss taken by end-2000). The capital injectionwill be provided by the government buying debenturesissued by the bank with a maturity of 10 years and paidfor with nontradable government bonds with matchingmaturity and carrying market-related interest rates. Therate on debentures is expected to be 1 percentage pointhigher than that carried by the government bond inorder to cover administrative costs for the government.

Box 18. Thailand: Capital Support Facilities

Appendix V


As of the end of 1998, the public sector contribu-tion for financial sector restructuring was close to

25 percent of GDP(Table 21). The largest portionof this, 20 percent of GDP, was used for li-quidity support; another 8 percent was used forrecapitalization.

101

Table 21. Thailand: Public Cost for Financial Sector Restructuring(As of the end of 1998)

Percent of GDP Billions of U.S. Dollars

Liquidity support assumed by the budget 15 20Recapitalization 8 11Purchases of nonperforming loans 0 0Interest cost (on budget) 2 3

Total 25 34


14 privatedomesticcommercialbanks(59%)

7 privatedomesticcommercialbanks1

(39%)

90 private financecompanies (20%)

1 majoritystate-owned

commercial bank (8%)

6 state-controlledcommercial banks (28%)

14 branches offoreign banks (6%)

14 branches offoreign banks (12%)

22 privatefinance companies (5%)

7 state-ownedspecialized banks (7%)

7 state-ownedspecialized banks (15%)

Financial sector at the end of 1996(share of assets)

Financial sector at the end of July 1999(share of assets)

1 public financecompany (0%)

1 public financecompany (1%)

Figure 11. Thailand: Progress in Financial Sector Restructuring

Source: National authorities; and IMF staff estimates.1 Two of these have greater than 50 percent foreign ownership.Note: As of end of July 1999, one commercial bank and 56 finance companies had been closed. The state had intervened in seven private, domestic

commercial banks and in 12 finance companies. Three intervened commercial banks and 12 finance companies had exited through merger. Categoriesare not mutually exclusive.

Alexander, William, Jeffrey Davis, Liam Patrick Ebrill,and Carl-Johan Lindgren, 1997, Systemic Bank Re-structuring and Macroeconomic Policy(Washington:International Monetary Fund).

Baliño, Tomás, and Angel Ubide, 1999, “The Korean Fi-nancial Crisis of 1997: A Strategy of Financial SectorReform,” IMF Working Paper 99/28 (Washington: In-ternational Monetary Fund).

Basel Committee on Banking Supervision, 1999, Supervi-sory Lessons to be Drawn from the Asian Crisis(Basel: Bank for International Settlements).

Daniel, James, Jeffrey Davis, and Andrew Wolfe, 1997,Fiscal Accounting of Bank Restructuring, IMF Paperon Policy Analysis and Assessment 97/5 (Washing-ton: International Monetary Fund).

Dattels, Peter, 1997, “Microstructure of Government Se-curities Markets,” in Coordinating Public Debt andMonetary Management,ed. byV. Sundararajan, P.Dattels, and H. Blommestein, (Washington: Interna-tional Monetary Fund).

Demirgüç-Kunt, A., and Enrica Detragiache, 1998, “TheDeterminants of Banking Crises in Developing andDeveloped Countries,” Staff Papers, InternationalMonetary Fund, Vol. 45, No. 1, pp.81–109.

Ding, Wei, Ilker Domaç, and Giovanni Ferri, 1998, “IsThere a Credit Crunch in East Asia?” World BankPolicy Research Working Paper No. 1959 (Washing-ton: World Bank).

Dollar, David, and Mary Hallward-Driemeier, 1998, “Cri-sis, Adjustment, and Reform in Thai Industry” (un-published; Washington: World Bank).

Downes, Patrick T., David Marston, and Inci Otker, 1999,Mapping Financial Sector Vulnerability in a Non-Crisis Country, IMF Policy Discussion Paper 99/4(Washington: International Monetary Fund).

Eichengreen, Barry, Donald Mathieson 1998, HedgeFunds and Financial Market Dynamics, IMF Occa-sional Paper No. 166 (Washington: InternationalMonetary Fund).

Enoch, Charles, and John Green, eds., 1997, BankingSoundness and Monetary Policy (Washington: Inter-national Monetary Fund).

Enoch, Charles, Gillian Garcia, and V. Sundararajan,1999, “Recapitalizing Banks with Public Funds: Se-lected Issues,” IMF Working Paper 99/139 (Washing-ton: International Monetary Fund).

Escolano, Julio, 1997, “Tax Treatment of Loan Losses ofBanks,” in Banking Soundness and Monetary Policy,ed. by Charles Enoch and John H. Green (Washing-ton: International Monetary Fund).

Feldstein, Martin, 1998, “Refocusing the IMF,” ForeignAffairs, Vol. 77, No. 2, (New York), pp. 20–33.

Folkerts-Landau, David, and Carl-Johan Lindgren, 1998,Toward a Framework for Financial Stability, WorldEconomic and Financial Surveys (Washington: Inter-national Monetary Fund).

Frécaut, Olivier, and Eric Sidgwick, 1998, “SystemicBanking Distress: The Need for an Enhanced Mone-tary Survey,” IMF Paper on Policy Analysis and As-sessment 98/9 (Washington: International MonetaryFund).

Furman, Jason, and Joseph Stiglitz, 1998, “EconomicCrises: Evidence and Insights from East Asia”,Brookings Papers on Economic Activity: 2, Brook-ings Institution.

García-Herrero, Alicia, 1997, “Monetary Impact of aBanking Crisis and the Conduct of Monetary Policy,”IMF Working Paper 97/124 (Washington: Interna-tional Monetary Fund).

Garcia, Gillian, 1999, “Deposit Insurance: A Survey ofActual and Best Practices in 1998,” IMF WorkingPaper 99/54 (Washington: International MonetaryFund).

Ghosh, Swati R., and Atish R. Ghosh, 1999, “East Asia inthe Aftermath: Was There a Crunch?” IMF WorkingPaper 99/38 (Washington: International MonetaryFund).

Goldstein, Morris, 1998, The Asian Financial Crisis:Causes, Cures and Systemic Implications (Washing-ton: Institute for International Economics).

Guitián, Manuel, 1997, “Bank Soundness: The Other Di-mension of Monetary Policy,” in Banking Soundnessand Monetary Policy, ed. by Enoch and Green(Washington: International Monetary Fund).

Hardy, Daniel, and Ceyla Pazarbasıoglu, 1998, “LeadingIndicators of Banking Crises: Was Asia Different?”IMF Working Paper 98/91 (Washington: InternationalMonetary Fund).

Ingves, Stefan, and Göran Lind, 1997, “Loan-Loss Recov-eries and Debt Resolution Agencies: The SwedishExperience,” in Bank Soundness and Monetary Pol-icy, ed. by Enoch and Green (Washington: Interna-tional Monetary Fund).

International Monetary Fund, 1997, World Economic Out-look, December 1997:Crisis in Asia, World Eco-nomic and Financial Surveys (Washington).

———, 1998, World Economic Outlook, October 1998:Financial Turbulence and the World Economy, WorldEconomic and Financial Surveys (Washington).

Bibliography

102

103

BIBLIOGRAPHY

Kaminsky, Graciela L., Saul Lizondo, and Carmen M.Reinhart, 1997, “Leading Indicators of CurrencyCrises,” IMF Working Paper 97/79 (Washington: In-ternational Monetary Fund).

Lane, Timothy, A. Ghosh, J. Hamann, S. Phillips, M.Schulze-Ghattas, and T. Tsikata, 1999, IMF-SupportedPrograms in Indonesia, Korea, and Thailand: A Pre-liminary Assessment, IMF Occasional Paper, No. 178(Washington: International Monetary Fund).

Leipziger, Danny M., 1997, Lessons from East Asia (AnnArbor: University of Michigan Press).

Lindgren, Carl-Johan, Gillian Garcia, and Matthew Saal,1996, Bank Soundness and Macroeconomic Policy(Washington: International Monetary Fund).

Lirio, Ricardo P., 1998, “The Central Bank and Non-BankFinancial Intermediaries” in Philippine Financial Al-manac 1997/98 (Manila).

Merrill Lynch, 1999, Asia-Pacific Banks: Progress and Is-sues in Bank Restructuring,Global Fundamental Eq-uity Research Department, February.

Nascimento, Jean-Claude, 1991, “The Crisis in the Finan-cial Sector and the Authorities’Reaction: The Philip-pines,” in Banking Crises: Cases and Issues,ed. by V.Sundararajan and Tomás J. T. Baliño (Washington:International Monetary Fund).

Nellor, David C. L., 1998, “Tax Policy and the Asian Cri-sis,” (Washington: International Monetary Fund).

Quintyn, Marc, 1997, “Government Securities VersusCentral Bank Securities in Developing Open MarketOperations,” in Coordinating Public Debt and Mone-tary Management,ed. by V. Sundararajan, P. Dattels,and H. Blommestein (Washington: InternationalMonetary Fund).

Radelet, Steven C., and Jeffrey Sachs, 1998, The Onsetof the East Asian Financial Crisis(Cambridge,Massachusetts: National Bureau of EconomicResearch).

Samsung Economic Research Institute (SERI), KoreanEconomic Trends, (Korea, various issues).

Sheng, Andrew, 1996, Bank Restructuring: Lessons fromthe 1980s(Washington: World Bank).

Sundararajan, V., and Tomás J.T. Baliño, 1991, BankingCrises: Cases and Issues(Washington: InternationalMonetary Fund).

———, 1990, Issues in Recent Banking Crises in Devel-oping Countries (Washington: International MonetaryFund).

Sundararajan, V., Dong He, and May Khamis, 1999,“Crises, Policy Responses and Steps to IncreaseTransparency—The Case of Asia,” Thomson BankWatch Conference on Banking Systems in Chaos(Washington: International Monetary Fund).

World Bank, 1998, East Asia: The Road to Recovery(Washington).

financial sector crisis and restructuringof capital flows.1 in turn, problems in financial...

Documents